\n",
"\n",
"This jupyter notebook shows an example of how to read in and use the data from simulations described in the paper https://arxiv.org/abs/1905.00910: \n",
"\n",
"**STROOPWAFEL: Simulating rare outcomes from astrophysical populations, with application to gravitational-wave sources**\n",
"\n",
"by Floor S. Broekgaarden, Stephen Justham, Selma E. de Mink, Jonathan Gair, Ilya Mandel, Simon Stevenson, Jim W. Barrett, Alejandro Vigna-Gómez and Coenraad J. Neijssel \n",
" \n",
"\n",
"The data can be found on Zenodo \n",
"This script is written for python 3, but should also work for python 2 by changing in the last function in the histogram function the 'density=True' to 'normed=True'\n",
"___\n",
"\n",
"Last updated September 2019 \n",
"For any queries, email: \n",
"\n",
"\n",
"__fsbroekgaarden@gmail.com__\n",
"\n",
"__fbroekgaarden@g.harvard.edu__"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Obtain data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- Download the file **STROOPWAFEL_data_BroekgaardenEtAll.tar.gz** from Zenodo and uncompress the data using: \n",
"```\n",
" tar xvzf STROOPWAFEL_data_BroekgaardenEtAll.tar.gz\n",
"```\n",
"\n",
"\n",
"The tar file contains the following 7 directories with data (DCO = Double Compact Object) : \n",
"\n",
"\n",
"**Traditional** # data of the simulations run with sampling from the birth distributions (traditional sampling) at Z = 0.001 metallicity \n",
"**ALL** \t # data of ALL DCOs (simulation 1) at Z = 0.001 metallicity with STROOPWAFEL sampling \n",
"**BHBH** \t # data of BH-BH DCOs (simulation 2) at Z = 0.001 metallicity with STROOPWAFEL sampling \n",
"**BHNS** \t # data of BH-NS DCOs (simulation 3) at Z = 0.001 metallicity with STROOPWAFEL sampling \n",
"**NSNS** \t # data of NS-NS DCOs (simulation 4) at Z = 0.001 metallicity with STROOPWAFEL sampling \n",
"**BHBH_50Msun** \t # data of BH-BH DCOs that have total mass <= 50 Msun (simulation 5) at Z = 0.001 metallicity with STROOPWAFEL sampling \n",
"**NSNS_50Myr** # data of NS-NS DCOs that merge within <= 50 Myrs (simulation 6) at Z = 0.001 metallicity with STROOPWAFEL sampling \n",
"\n",
"which correspond with the 6 STROOPWAFEL sampled simulations + the reference \"birth distribution Monte Carlo\" sampled simulation (\"Traditional\"). These simulations are summarized in Table 2 of https://arxiv.org/pdf/1905.00910.pdf "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Analyze STROOPWAFEL data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Using the functions **obtaindataSTROOPWAFEL** and **maskTargetDCOsSTROOPWAFEL** which are defined below we can obtain the needed parameters from the data and, for example, calculate the merger rate of all DCOs for both the traditional and STROOPWAFEL sampled simulations. \n",
"\n",
"- the function **obtaindataSTROOPWAFEL** obtains the wanted parameter that are described in README_DCOs.txt and README_systems.txt, such as the DCO masses or separation. \n",
"- the function **maskTargetDCOsSTROOPWAFEL** returns a mask that can filter the wanted DCOtype. An example is BH-BH binaries that merge in a Hubble time. \n",
"\n",
"\n",
"In the remaining notebook we will provide examples. For these examples we use the simulations from Traditional sampling (simulation 0) and ALL DCOs (simulation 1). But these can easily be changed to any of 'Traditional', 'ALL', 'BHNS', 'BBH', 'NSNS', 'NSNS_50Myr' or 'BHBH_50Msun' corresponding with the 6 simulations performed in our study. To do this, change the parameter ```simulation``` to the wanted simulation from the list 'Traditional', 'ALL', 'BHNS', 'BBH', 'NSNS', 'NSNS_50Myr' or 'BHBH_50Msun' and changing, if needed, ```DCOtype``` to the corresponding target binary type from the list ['ALL', 'BHNS', 'BBH', 'NSNS' or 'NSNS_50Myr'] \n",
"\n",
"[1. Calculate speed up STROOPWAFEL sampling](1.-Calculate-speed-up-STROOPWAFEL-sampling) \n",
"[2. Calculate merger rate in simulation](#2.-Calculate-merger-rate-in-simulation) \n",
"[3. Plot fractional statistical sampling uncertainties](#3.-Plot-fractional-statistical-sampling-uncertainties) \n",
"[4. Plot chirp mass distribution](#4.-Calculate-merger-rate-in-simulation) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### import libraries and add path to data. "
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd # can be downloaded here: https://pandas.pydata.org/\n",
"import numpy as np # can be installed through https://docs.scipy.org/doc/numpy/user/install.html\n",
"\n",
"# path to directory with Traditional and STROOPWAFEL directories 'ALL', 'BHBH' etc.. \n",
"pathData = '/Users/floorbroekgaarden/Programs/COMPAS/popsynth/Papers/BroekgaardenEtAl/STROOPWAFELmethod/PublicDATA/STROOPWAFEL_data_BroekgaardenEtAll'\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"def obtaindataSTROOPWAFEL(simulation='ALL', datafile='allDoubleCompactObjects.dat', path=pathData, paramHeader='separationInitial'):\n",
" '''\n",
" returns the wanted parameter and corresponding weights for a given simulation. In case of the STROOPWAFEL simulation\n",
" the function combines the data from the exploration and refinement phase\n",
" //\n",
" parameters:\n",
" simulation # simulation name. options are 'Traditional', 'ALL', 'BHNS', 'BHBH', 'NSNS', 'NSNS_50Myr' and 'BHBH_50Msun'\n",
" datafile # datafile that contains wanted parameter; either 'allDoubleCompactObjects.dat' or 'allSystems.dat'\n",
" path # path to the directory with all the data\n",
" paramHeader # header name of the wanted parameter in the allSystems.dat or allDoubleCompactObjects.dat files\n",
" '''\n",
" \n",
" # read in datafiles. \n",
" # stroopwafel sampling has 2 directores: exploration and refinement phase\n",
" if (simulation=='Traditional') :\n",
" data = pd.read_csv(path + '/' + simulation + '/' + datafile, sep='\\t', header=2)\n",
" weights = np.ones(len(data['ID'])) # all traditional weights are 1\n",
"\n",
" else: # Stroopwafel sampling\n",
" path_expl = path + '/' + simulation + '/exploration/'\n",
" path_ref = path + '/' + simulation + '/refinement/'\n",
"\n",
" # Set name of weights textfile\n",
" if datafile=='allDoubleCompactObjects.dat':\n",
" weightsfilename = 'allDoubleCompactObjectsWeights.txt'\n",
" elif datafile=='allSystems.dat':\n",
" weightsfilename = 'allSystemsWeights.txt'\n",
" else:\n",
" print('error: name of datafile not correct in obtaindataSTROOPWAFEL()') \n",
" \n",
" # read in datafiles\n",
" data_expl = pd.read_csv(path_expl + datafile, sep='\\t', header=2)\n",
" data_weights_expl = pd.read_csv(path_expl + weightsfilename, sep='\\t', header=2)\n",
"\n",
" data_ref = pd.read_csv(path_ref + datafile, sep='\\t', header=2)\n",
" data_weights_ref = pd.read_csv(path_ref + weightsfilename, sep='\\t', header=2)\n",
"\n",
" # combine data and weights from exploration and refinement \n",
" data = pd.concat([data_expl, data_ref])\n",
" weights = pd.concat([data_weights_expl['weight'], data_weights_ref['weight']])\n",
" \n",
" param = data[paramHeader]\n",
" \n",
" return np.asarray(param), np.asarray(weights)\n",
"\n",
"\n",
"def maskTargetDCOsSTROOPWAFEL(DCOtype='ALL', simulation='ALL', boolDCOmask=[1,1,0], path=pathData):\n",
" \"\"\"\n",
" returns mask of DCOs of interest\n",
" //\n",
" parameters:\n",
" DCOtype # describes target binary. Options are: 'ALL', 'BHNS', 'BBH', 'NSNS', 'NSNS_50Myr' and 'BHBH_50Msun\n",
" simulation # name of simulation. Options are: Traditional, 'ALL', 'BHNS', 'BBH', 'NSNS', 'NSNS_50Myr' and 'BHBH_50Msun'\n",
" boolDCOmask = [Hubble, RLOF, Pessimistic] # boolean values whether to mask mergers in a HUbble time, \n",
" binaries that have RLOFSecondaryAfterCEE = True, and Pessimistic binaries (i.e. optimisticCEFlag == 0)\n",
" path # pathname to Directory where _exploratory & _refinement directories are\n",
" \"\"\"\n",
" \n",
" Hubble, RLOF, Pessimistic = boolDCOmask\n",
" \n",
" # all needed parameters are in the DCO file\n",
" datafile = 'allDoubleCompactObjects.dat'\n",
" \n",
" # data for DCO type mask\n",
" stellarType1, _ = obtaindataSTROOPWAFEL(simulation, datafile, path, paramHeader='stellarType1') \n",
" stellarType2, _ = obtaindataSTROOPWAFEL(simulation, datafile, path, paramHeader='stellarType2') \n",
" # data for boolDCOmask: \n",
" mergesInHubbleTimeFlag, _ = obtaindataSTROOPWAFEL(simulation, datafile, path, paramHeader='mergesInHubbleTimeFlag') \n",
" RLOFSecondaryAfterCEE, _ = obtaindataSTROOPWAFEL(simulation, datafile, path, paramHeader='RLOFSecondaryAfterCEE') \n",
" optimisticCEFlag, _ = obtaindataSTROOPWAFEL(simulation, datafile, path, paramHeader='optimisticCEFlag') \n",
" \n",
" if (DCOtype == 'NSNS_50Myr') : # obtain merger time\n",
" tc, _ = obtaindataSTROOPWAFEL(simulation, datafile, path, paramHeader='tc') \n",
" if (DCOtype == 'BHBH_50Msun') : # obtain BH masses \n",
" M1, _ = obtaindataSTROOPWAFEL(simulation, datafile, path, paramHeader='M1') \n",
" M2, _ = obtaindataSTROOPWAFEL(simulation, datafile, path, paramHeader='M2') \n",
" Mtot = M1 + M2 \n",
" \n",
" # mask binaries of given simulation (i.e. DCO type)\n",
" if (DCOtype == 'all') | (DCOtype == 'ALL') :\n",
" mask0 = ((stellarType1== 14) | (stellarType1== 13))\n",
" elif (DCOtype == 'BBH') | (DCOtype == 'BHBH') :\n",
" mask0 = ((stellarType1== 14) & (stellarType2== 14))\n",
" elif (DCOtype == 'BNS') | (DCOtype == 'NSNS') :\n",
" mask0 = ((stellarType1 == 13) & (stellarType2 == 13))\n",
" elif (DCOtype == 'BHNS') | (DCOtype == 'NSBH'):\n",
" mask0 = ((stellarType1== 13) & (stellarType2== 14)) | \\\n",
" ((stellarType1== 14) & (stellarType2== 13) ) \n",
" elif (DCOtype == 'NSNS_50Myr') :\n",
" mask0 = ((stellarType1 == 13) & (stellarType2 == 13) & (tc <= 50) )\n",
" elif (DCOtype == 'BHBH_50Msun'):\n",
" mask0 = ((stellarType1== 14) & (stellarType2== 14) & (Mtot >= 50)) \n",
" else:\n",
" print('error: DCO type not known')\n",
" \n",
" # boolDCOmasks:\n",
" # Hubble mask\n",
" if Hubble:\n",
" mask1 = (mergesInHubbleTimeFlag==True) \n",
" elif not Hubble:\n",
" mask1 = (mergesInHubbleTimeFlag==True) | (mergesInHubbleTimeFlag==False) \n",
" # RLOF mask\n",
" if RLOF:\n",
" mask2 = (RLOFSecondaryAfterCEE==False)\n",
" elif not RLOF:\n",
" mask2 = (RLOFSecondaryAfterCEE==False) | (RLOFSecondaryAfterCEE==True)\n",
" # Pessimistic mask : if True mask systems that have optimistic CE flag ==1\n",
" if Pessimistic:\n",
" mask3 = np.logical_not(optimisticCEFlag==1)\n",
" elif not Pessimistic:\n",
" mask3 = np.logical_not(optimisticCEFlag==1) + \\\n",
" np.logical_not(optimisticCEFlag==0) \n",
" \n",
" # combine the different masks \n",
" combinedmask = mask0 * mask1 * mask2 * mask3\n",
"\n",
" return combinedmask\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 1. Calculate speed up STROOPWAFEL sampling"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can now calculate the speed up from STROOPWAFEL sampling compared to Traditional sampling by using the functions above to obtain the number of binaries of the target simulation that we find in each simulation. "
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"The target binaries are: ALL DCOs that merge in a Hubble time\n",
"The number of target binaries found in the Traditional simulation equals 6711\n",
"\n"
]
}
],
"source": [
"\n",
"#Traditional sampling (simulation 0)\n",
"sim = 'Traditional'\n",
"# obtain the separations and weights of all DCOs from Traditional simulation\n",
"separationInitial_0, _ = obtaindataSTROOPWAFEL(paramHeader='separationInitial', simulation=sim,\\\n",
" datafile='allDoubleCompactObjects.dat', path=pathData)\n",
"# create mask with ALL DCOs that merge in a Hubble time (boolDCOmask=[1,1,0])\n",
"DCOmask_0 = maskTargetDCOsSTROOPWAFEL(DCOtype='ALL', simulation=sim, boolDCOmask=[1,1,0], path=pathData)\n",
"\n",
"# the number of target binaries in this simulation is\n",
"NDCO_0 = len(separationInitial_0[DCOmask_0]) \n",
"print()\n",
"print('The target binaries are: ', 'ALL DCOs that merge in a Hubble time')\n",
"print('The number of target binaries found in the', sim, 'simulation equals ', NDCO_0)\n",
"print()\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"The target binaries are: ALL DCOs that merge in a Hubble time\n",
"The number of target binaries found in the ALL STROOPWAFEL simulation equals 234768\n",
"\n"
]
}
],
"source": [
"#STROOPWAFEL sampling focused on ALL DCOs (simulation 1)\n",
"sim = 'ALL'\n",
"# obtain the separations and weights of all DCOs from Traditional simulation\n",
"separationInitial_1, _ = obtaindataSTROOPWAFEL(paramHeader='separationInitial', simulation=sim,\\\n",
" datafile='allDoubleCompactObjects.dat', path=pathData)\n",
"# create mask with ALL DCOs that merge in a Hubble time (boolDCOmask=[1,1,0])\n",
"DCOmask_1 = maskTargetDCOsSTROOPWAFEL(DCOtype='ALL', simulation=sim, boolDCOmask=[1,1,0], path=pathData)\n",
"\n",
"# the number of target binaries in this simulation is\n",
"NDCO_1 = len(separationInitial_1[DCOmask_1]) \n",
"print()\n",
"print('The target binaries are: ', sim, ' DCOs that merge in a Hubble time')\n",
"print('The number of target binaries found in the', sim, 'STROOPWAFEL simulation equals ', NDCO_1)\n",
"print()\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"the gain = 34.98256593652213\n"
]
}
],
"source": [
"print('the gain = ', float(NDCO_1)/NDCO_0)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"So in the end there are **6711** DCOs that merge in a Hubble time in the Traditional sampled simulation, and **234768** target binaries in the STROOPWAFEL sampled simulation. These are the numbers that are quoted in Table 1 of https://arxiv.org/pdf/1905.00910.pdf where **$N_{T,traditional}$** are the number of target binaries from the Traditional sampled simulation (here 6711 for ALL mergers) and **$N_{T,STROOPWAFEL}$** are the Number of binaries from the STROOPWAFEL simulation (here 234768). These numbers are also quoted in parentheses in Fig. 6 and Fig. 7 of https://arxiv.org/pdf/1905.00910.pdf. \n",
"\n",
"The overall gain in DCOs of the target population is given by: \n",
"\n",
"$ N_{T,STROOPWAFEL} / N_{T,traditional} = 234768 / 6711 \\approx 35 $ \n",
"\n",
"which is the number corresponding to the results for the **'ALL DCO mergers'** simulation shown in the last column of Table 1 (first row) and top left panel of Fig. 4 of https://arxiv.org/pdf/1905.00910.pdf\n",
"\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**with the function below we can obtain all the overall gains quoted in the last column of Table 2 of https://arxiv.org/pdf/1905.00910.pdf**"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"def calculateGainSTROOPWAFEL(simulation='ALL', DCOtype='ALL', path=pathData):\n",
" \"\"\"function that calculates the gain in Table 2 of the STROOPWAFEL paper\n",
" //\n",
" parameters:\n",
" DCOtype # describes target binary. Options are: 'ALL', 'BHNS', 'BBH', 'NSNS', 'NSNS_50Myr' and 'BHBH_50Msun\n",
" simulation # name of simulation. Options are: Traditional, 'ALL', 'BHNS', 'BBH', 'NSNS', 'NSNS_50Myr' and 'BHBH_50Msun'\n",
" path # pathname to Directory where _exploratory & _refinement directories are\n",
" \"\"\"\n",
" \n",
" ####Traditional sampling (simulation 0)\n",
" # obtain the separations of all DCOs from Traditional simulation\n",
" separationInitial_0, _ = obtaindataSTROOPWAFEL(paramHeader='separationInitial', simulation='Traditional',\\\n",
" datafile='allDoubleCompactObjects.dat', path=path)\n",
" # create mask with ALL DCOs that merge in a Hubble time (boolDCOmask=[1,1,0])\n",
" DCOmask_0 = maskTargetDCOsSTROOPWAFEL(DCOtype=DCOtype, simulation='Traditional', boolDCOmask=[1,1,0], path=path)\n",
" # the number of target binaries in the Traditional simulation is\n",
" NDCO_0 = len(separationInitial_0[DCOmask_0]) \n",
"\n",
" \n",
" ## STROOPWAFEL sampling ###\n",
" # obtain the separations and weights of all DCOs from Traditional simulation\n",
" separationInitial_1, _ = obtaindataSTROOPWAFEL(paramHeader='separationInitial', simulation=simulation,\\\n",
" datafile='allDoubleCompactObjects.dat', path=path)\n",
" # create mask with ALL DCOs that merge in a Hubble time (boolDCOmask=[1,1,0])\n",
" DCOmask_1 = maskTargetDCOsSTROOPWAFEL(DCOtype=DCOtype, simulation=simulation, boolDCOmask=[1,1,0], path=path)\n",
" # the number of target binaries in the STROOPWAFEL simulation is\n",
" NDCO_1 = len(separationInitial_1[DCOmask_1]) \n",
" \n",
" \n",
" gain = int(round(float(NDCO_1) / NDCO_0))\n",
" \n",
" return gain \n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"gain of simulation 1 (ALL mergers) = 35\n",
"gain of simulation 2 (BHBH mergers) = 53\n",
"gain of simulation 3 (BHNS mergers) = 39\n",
"gain of simulation 4 (NSNS mergers) = 45\n",
"gain of simulation 5 (BHBH with Mtot >=50 Msun mergers) = 203\n",
"gain of simulation 6 (NSNS with tcoalescence <= 50 Myr mergers) = 24\n"
]
}
],
"source": [
"## print the gains of STROOPWAFEL from Table 2:\n",
"print( 'gain of simulation 1 (ALL mergers) = ', calculateGainSTROOPWAFEL(simulation='ALL', DCOtype='ALL', path=pathData))\n",
"print( 'gain of simulation 2 (BHBH mergers) = ', calculateGainSTROOPWAFEL(simulation='BHBH',DCOtype='BHBH', path=pathData))\n",
"print( 'gain of simulation 3 (BHNS mergers) = ', calculateGainSTROOPWAFEL(simulation='BHNS',DCOtype='BHNS', path=pathData))\n",
"print( 'gain of simulation 4 (NSNS mergers) = ', calculateGainSTROOPWAFEL(simulation='NSNS',DCOtype='NSNS', path=pathData))\n",
"print( 'gain of simulation 5 (BHBH with Mtot >=50 Msun mergers) = ', calculateGainSTROOPWAFEL(simulation='BHBH_50Msun',DCOtype='BHBH_50Msun', path=pathData))\n",
"print( 'gain of simulation 6 (NSNS with tcoalescence <= 50 Myr mergers) = ', calculateGainSTROOPWAFEL(simulation='NSNS_50Myr',DCOtype='NSNS_50Myr', path=pathData))\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 2. Calculate merger rate in simulation"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"It is also possible to calculate the fractional merger rate of the target population in the COMPAS simulation. This is done similar to the example above, but by incorporating the weights. The function\n",
"`calculateFormationRate()` below calculates the ractional merger rate for a given simulation and target population"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"def printDCOmaskString(boolDCOmask=[1,1,0]):\n",
" \"\"\"\n",
" returns string that contains the boolDCOmask selection, \n",
" This is printed in the function calculateFormationRate()\n",
" //\n",
" parameters:\n",
" boolDCOmask = [Hubble, RLOF, Pessimistic] # boolean values whether to mask mergers in a HUbble time, \n",
" binaries that have RLOFSecondaryAfterCEE = True, and Pessimistic binaries (i.e. optimisticCEFlag == 0)\n",
" \"\"\"\n",
" Hubble, RLOF, Pessimistic = boolDCOmask\n",
" \n",
" # boolDCOmasks:\n",
" # Hubble mask\n",
" if Hubble:\n",
" string1 = 'mergers in a Hubble time, ' \n",
" elif not Hubble:\n",
" string1 = ''\n",
" # RLOF mask\n",
" if RLOF:\n",
" string2 = ' that do not have Roche Lobe overflow immediately after the Common Envelope event,'\n",
" elif not RLOF:\n",
" string2 = ''\n",
" # Pessimistic mask : if True mask systems that have optimistic CE flag ==1\n",
" if Pessimistic:\n",
" string3 = ' which are optimistic.' \n",
" elif not Pessimistic:\n",
" string3 = ' which are pessimistic.'\n",
" \n",
" return string1 + string2 + string3\n",
"\n",
"\n",
"def calculateFormationRate(simulation='ALL', DCOtype='ALL', boolDCOmask=[1,1,0], path=pathData):\n",
" \"\"\"\n",
" returns fractional formation rate of the target binary (given by DCOtype and boolDCOmask)\n",
" //\n",
" parameters:\n",
" simulation # name of simulation. Options are: Traditional, 'ALL', 'BHNS', 'BBH', 'NSNS', 'NSNS_50Myr' and 'BHBH_50Msun'\n",
" DCOtype # describes target binary. Options are: 'ALL', 'BHNS', 'BBH', 'NSNS', 'NSNS_50Myr' and 'BHBH_50Msun\n",
" boolDCOmask = [Hubble, RLOF, Pessimistic] # boolean values whether to mask mergers in a HUbble time, \n",
" binaries that have RLOFSecondaryAfterCEE = True, and Pessimistic binaries (i.e. optimisticCEFlag == 0)\n",
" path # pathname to Directory where _exploratory & _refinement directories are\n",
" \"\"\"\n",
" # obtain the separations and weights of all DCOs from Traditional simulation from allDoubleCompactObjects.dat\n",
" _, weights = obtaindataSTROOPWAFEL(paramHeader='separationInitial', simulation=simulation,\\\n",
" datafile='allDoubleCompactObjects.dat', path=path)\n",
" # create mask with ALL DCOs that merge in a Hubble time (boolDCOmask=[1,1,0])\n",
" DCOmask = maskTargetDCOsSTROOPWAFEL(DCOtype=DCOtype, simulation=simulation,\\\n",
" boolDCOmask=boolDCOmask, path=path)\n",
"\n",
" # obtain a parameter from the allSystems.dat to get the total Number of binaries simulated of Traditional simulation\n",
" ID, _ = obtaindataSTROOPWAFEL(paramHeader='ID', simulation=simulation,\\\n",
" datafile='allSystems.dat', path=path)\n",
" NbinariesSampled = len(ID)\n",
" \n",
" # the total weight of target binaries in this simulation is\n",
" totalweights_DCO = np.sum(weights[DCOmask])\n",
" \n",
" # print the outcomes\n",
" DCOmaskstring = printDCOmaskString(boolDCOmask=boolDCOmask)\n",
" print('This is a calculation for the %s simulation'%sim)\n",
" print('The target binaries are: ', DCOtype, DCOmaskstring)\n",
" print('The total weight of the target binaries in the', simulation, 'simulation with ', \\\n",
" NbinariesSampled ,' binaries equals ', totalweights_DCO)\n",
" print('the estimated fractional rate of the target population in our simulation equals ', totalweights_DCO/NbinariesSampled)\n",
" \n",
" return\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Example: fractional merger rate of ALL mergers from the Traditional sampled simulation (simulation 0)"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"This is a calculation for the ALL simulation\n",
"The target binaries are: ALL mergers in a Hubble time, that do not have Roche Lobe overflow immediately after the Common Envelope event, which are pessimistic.\n",
"The total weight of the target binaries in the Traditional simulation with 1000000 binaries equals 6711.0\n",
"the estimated fractional rate of the target population in our simulation equals 0.006711\n"
]
}
],
"source": [
"calculateFormationRate(simulation = 'Traditional', DCOtype='ALL', boolDCOmask=[1,1,0], path=pathData)\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Example: fractional merger rate of ALL mergers from the STROOPWAFEL sampled simulation (simulation 1)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"This is a calculation for the ALL simulation\n",
"The target binaries are: ALL mergers in a Hubble time, that do not have Roche Lobe overflow immediately after the Common Envelope event, which are pessimistic.\n",
"The total weight of the target binaries in the ALL simulation with 1000001 binaries equals 6562.143916108497\n",
"the estimated fractional rate of the target population in our simulation equals 0.006562137353971143\n"
]
}
],
"source": [
"calculateFormationRate(simulation='ALL', DCOtype='ALL', boolDCOmask=[1,1,0], path=pathData)\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The estimated fractional rates of the target population, $\\mathcal{R_{\\rm{T}}}, $ of the target population should be similar between the Traditional sampled simulation (simulation 0) and the STROOPWAFEL sampled simulation (here simulation 1). Small differences between the two estimates are caused by sampling (Poisson) noise. These estimated physical rates of the target population (i.e. $\\approx 0.0066$ for ALL mergers in a Hubble time) are shown in Fig. 2 and Fig. 3 of https://arxiv.org/pdf/1905.00910.pdf with coloured vertical bars. "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 3. Plot fractional statistical sampling uncertainties"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Using the data from all the simulations, the function `plotUncertainties()` below recreates Figure 5 in https://arxiv.org/pdf/1905.00910.pdf. \n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [],
"source": [
"import matplotlib.pyplot as plt\n",
"from matplotlib import rc \n",
"from matplotlib import rcParams\n",
"\n",
"# below is just for the text\n",
"rc('font', family='serif', weight = 'bold')\n",
"rc('text', usetex=True)\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [],
"source": [
"\n",
"\n",
"def plotUncertainties(ax=None, DCOtype='ALL', i=0, path=pathData): \n",
" \"\"\"\n",
" (1) plots the fractional statistical sampling uncertainties for the given target population\n",
" for the Traditional and STROOPWAFEL sampled simulation. (2) prints the ratio between the two fractional\n",
" uncertainties. \n",
" //\n",
" parameters:\n",
" DCOtype # describes target binary. Options are: 'ALL', 'BHNS', 'BBH', 'NSNS', 'NSNS_50Myr' and 'BHBH_50Msun\n",
" i # indicator for nr of simulation used to determine labels and colors\n",
" path # pathname to Directory where _exploratory & _refinement directories are\n",
" \"\"\"\n",
" \n",
" # list of rounded factors of improvement in fractional uncertainty by STROOPWAFEL\n",
" factorlist = [r'\\textbf{4}',r'\\textbf{6}',r'\\textbf{4}',r'\\textbf{4}',r'\\textbf{10.5}',r'\\textbf{3}']\n",
"\n",
"\n",
" # Traditional sampled simulation\n",
" # obtain total number of binaries sampled. this is 1E6\n",
" ID, _ = obtaindataSTROOPWAFEL(paramHeader='ID', simulation='Traditional',\\\n",
" datafile='allSystems.dat', path=path)\n",
" NbinariesSampled_0 = len(ID)\n",
" # obtain the weights\n",
" _, weights_0 = obtaindataSTROOPWAFEL(paramHeader='M2', simulation='Traditional',\\\n",
" datafile='allDoubleCompactObjects.dat', path=path) \n",
" DCOmask_0 = maskTargetDCOsSTROOPWAFEL(DCOtype=DCOtype, simulation='Traditional', boolDCOmask=[1,1,0], path=path) \n",
" weights_0 = weights_0[DCOmask_0]\n",
" \n",
" # estimate rate of DCOtype (cf. Eq. 14 of https://arxiv.org/pdf/1905.00910.pdf)\n",
" rate_0 = np.sum(weights_0)/NbinariesSampled_0\n",
" \n",
" \n",
" # STROOPWAFEL sampled simulation\n",
" # obtain total number of binaries sampled. this is ~ 1E6\n",
" ID, _ = obtaindataSTROOPWAFEL(paramHeader='ID', simulation=DCOtype,\\\n",
" datafile='allSystems.dat', path=path)\n",
" NbinariesSampled_1 = len(ID)\n",
" # obtain the weights\n",
" _, weights_1 = obtaindataSTROOPWAFEL(paramHeader='M2', simulation=DCOtype,\\\n",
" datafile='allDoubleCompactObjects.dat', path=path) \n",
" DCOmask_1 = maskTargetDCOsSTROOPWAFEL(DCOtype=DCOtype, simulation=DCOtype, boolDCOmask=[1,1,0], path=path) \n",
" weights_1 = weights_1[DCOmask_1] \n",
" \n",
" # estimate rate of DCOtype (cf. Eq. 14 of https://arxiv.org/pdf/1905.00910.pdf)\n",
" rate_1 = np.sum(weights_1)/NbinariesSampled_1\n",
" \n",
" # samples that are not masked by DCOmask have outcome '0' (phi(x_i) = 0)\n",
" Nzeros_0 = int(NbinariesSampled_0 - len(weights_0))\n",
" Nzeros_1 = int(NbinariesSampled_1 - len(weights_1))\n",
" \n",
" # get all phi_x outcomes and estimate fractional uncertainty for Traditional simulation\n",
" phi_x_0 = np.concatenate((np.zeros(Nzeros_0), weights_0)) # phi(x) in Eq. 14\n",
" uncertainty_0 = np.std(phi_x_0, ddof = 1) / np.sqrt(NbinariesSampled_0)\n",
" fractional_uncertainty_0 = uncertainty_0 / rate_0 # Eq. 15\n",
" \n",
" # get all phi_x outcomes and estimate fractional uncertainty for STROOPWAFEL simulation\n",
" phi_x_1 = np.concatenate((np.zeros(Nzeros_1), weights_1)) # phi(x) in Eq. 14\n",
" uncertainty_1 = np.std(phi_x_1, ddof = 1) / np.sqrt(NbinariesSampled_1)\n",
" fractional_uncertainty_1 = uncertainty_1 / rate_1 # Eq. 15\n",
" \n",
"\n",
" ax.bar(i-0.14, fractional_uncertainty_1, lw = 2, color=colors[i], zorder = 100, width = 0.5, alpha = 0.85, linewidth = 0, edgecolor = colors[i])\n",
" ax.bar(i+0.14, fractional_uncertainty_0, lw = 2, color='gray', zorder = 2, width = 0.5, alpha = 0.85, edgecolor = 'gray', linewidth = 0, hatch=\"/\")\n",
" ax.text(i+0.14, fractional_uncertainty_0 + 0.0012, factorlist[i] +r'$\\mathbf{\\times}$', fontsize=40, weight='bold', horizontalalignment='center', verticalalignment='center', color = colors[i])\n",
" ax.tick_params(labelsize=30)\n",
"\n",
"\n",
" print('\\n' + 'For the target population = ', DCOtype)\n",
" print('STROOPWAFEL sampling results in ', round(fractional_uncertainty_0/fractional_uncertainty_1, 2)\\\n",
" , 'times smaller fractional sampling uncertainties')\n",
"\n",
" return\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/ipykernel_launcher.py:61: MatplotlibDeprecationWarning: Saw kwargs ['lw', 'linewidth'] which are all aliases for 'linewidth'. Kept value from 'linewidth'. Passing multiple aliases for the same property will raise a TypeError in 3.3.\n",
"/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/ipykernel_launcher.py:62: MatplotlibDeprecationWarning: Saw kwargs ['lw', 'linewidth'] which are all aliases for 'linewidth'. Kept value from 'linewidth'. Passing multiple aliases for the same property will raise a TypeError in 3.3.\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"For the target population = ALL\n",
"STROOPWAFEL sampling results in 3.87 times smaller fractional sampling uncertainties\n",
"\n",
"For the target population = BHBH\n",
"STROOPWAFEL sampling results in 6.29 times smaller fractional sampling uncertainties\n",
"\n",
"For the target population = BHNS\n",
"STROOPWAFEL sampling results in 3.92 times smaller fractional sampling uncertainties\n",
"\n",
"For the target population = NSNS\n",
"STROOPWAFEL sampling results in 4.07 times smaller fractional sampling uncertainties\n",
"\n",
"For the target population = BHBH_50Msun\n",
"STROOPWAFEL sampling results in 10.44 times smaller fractional sampling uncertainties\n",
"\n",
"For the target population = NSNS_50Myr\n",
"STROOPWAFEL sampling results in 3.07 times smaller fractional sampling uncertainties\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAABHgAAAR4CAYAAAB98mFDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzde3DcZ53n+8+jxErAstJWgJia2CStzFYlXBJachLCMifBLQzM7MwsSHbN7KmlZoK7GSa1BSGjRodLFhgQrSFAbaWWURvYZauWM7aanMMZIBi1Q2qGhEusJgHi1DlE7SEOhR2w3JblLJYGPecP9a/dkvp+f7rfryqVrdbv932e7t/PiX8fPxdjrRUAAAAAAADc1dPqDgAAAAAAAKA2BDwAAAAAAACOI+ABAAAAAABwHAEPAAAAAACA4wh4AAAAAAAAHHd5qzvQKV72spfZ6667rtXdcM6FCxe0devWVncDqBj3LlzEfQtXce/CRdy3cBX3bvubm5v7jbX25RtfJ+Cpk+uuu07Hjh1rdTec8+ijj+rOO+9sdTeAinHvwkXct3AV9y5cxH0LV3Hvtj9jzC/yvc4ULQAAAAAAAMcR8AAAAAAAADiOgAcAAAAAAMBxBDwAAAAAAACOI+ABAAAAAABwHAEPAAAAAACA4wh4AAAAAAAAHEfAAwAAAAAA4DgCHgAAAAAAAMcR8AAAAAAAADiOgAcAAAAAAMBxBDwAAAAAAACOu7zVHQAAAAAAAO3n4tKynvnOz/X8k7/S+dNLWjy1pN6tverf0af+a/q0bUefbtr7++rfsa3VXYUIeAAAAAAAwAY/+EpSTz10XJL0ezfv0PV37NKV267Qb89f1InHn1Pq8eckSU89dFz+O3bptne9nqCnxQh4AAAAAABA1jc+mtAvnzol/x279Ad/fbuu6Otd9/Pb3xXQbPSfsiFP6vHn9PxTp/TOz76NkKeFWIMHAAAAAABIktI/WdIvnzolSfpNakG/fvZM3uNue9fr132/fGFZs1P/3PD+oTACHgAAAAAAIElK//RC9veLp5b0zfuP5j2uf8c29W5dP7LnN/MLuri03ND+oTACHgAAAAAAIElavbi66bVfz+cfxZPPxaWL9ewOKsAaPA6w1urixYtaXFzU0tKSVlZWtLq6+Q+di6666io988wzre4GUDHuXbioXvdtT0+PtmzZor6+PvX39+uKK66QMaYOPQQAAK12xSt6tXLuf6177eWDV+c9dvnC5tE65azBMxv9J90y+uqCdcuReuwXmv/eLzQS+YOqa3QaAp42t7KyopMnT2p1dVX9/f165Stfqd7eXvX09HTEX6bPnz+vbdtYhAvu4d6Fi+px31prtbq6quXlZZ0/f17PP/+8enp6tHPnTm3ZsqVOPQUAAK2yI7hd6dWX6DfzC+rd2qv/7a9vy3vc4qnzm177vZt3lNXGb1ILeujeh/WHH9uja295ZcV9TD32C81O/bP6d/RVfG4nI+BpYysrK3ruuefk8/k0MDDQEYEOAMBtxhhddtlleslLXqKXvOQlevnLX66FhQU999xz2rVrFyEPAACOu+zKHr3zs2/XxaXlTbtn5Tp+5OebXvuD9+YPgzZ6xwNv10Mf+Ja+ef/RikOe55/8VTbceccDby/7vG7AGjxtylqrkydPyufz6eqrrybcAQC0JWOMrr76avl8Pp08eVLW2lZ3CQAA1EGxcOfX82f01EPH1702Mv6msrdIv6KvV3/4sT3q3dqrb95/tOw1fp5/8lf65v1Hs+FOsT52IwKeNnXx4kWtrq5qYGCg1V0BAKCkgYEBra6u6uJFFlYEAKBTLZ46r+NHfq6H7n04+9rv3bxDfzb9J/K/8VUV1erfsU3v/Ozb1Lu1Vw/d+3DJkIdwpzSmaLWpxcVF9ff3M3IHAOAEY4z6+/u1uLioK6+8stXdAQAAdXL8yM/1w6/8eNOCyv07+hQcf1NNCyV7Ic/X7n1YD937sN7x2bflrffr+TP65v1H1bt1beQP4U5+jOBpU0tLSyzgCgBwyrZt27S0tNTqbgAAgDq69uYduu1dr9eb3nub3vTe23TzO25S/44+LZ5a0kP3Pqyv3fstpR77RdX1c0fyfOMjm6dr/Xr+jB6692H1bu3VOz/7trKngXUjAp42tbKyot5eUkkAgDt6e3u1srLS6m4AAIA66t+xTTft/f3s1+3vCujPpv9UN+69QZL0m/kFzU79s77x0YQuLm3eNr3cNv7oE3skaV3IQ7hTGQKeNrW6uqqeHi4PAMAdPT09Wl1dbXU3AABAE/zBe29ft035L586pYc+8K2qQ56XD16tP/rEHi1fWNY3PnJUqcd+oW985CjhTgVIENoY6+8AAFzC/7cAAOgu19+xa933i6eW9MP/kay63ssHr9Y7Pvs2LV9Y1uzUP0sS4U4FCHgAAAAAAEDFrr35lZtee+bIs1WP4vH0bmW5kmoQ8AAAAAAAAEnSb08v6/iRn5e1cHLuFK1cv3zqV1W1vXjqvL7xkaOSpDe99zYtX1jW1+59WIunzldVr9uwTToAAAAAANCTDz2tX3z1Bf1CL0iSXjb4tN752bdXXOe3VYzgWTx1Xl+792EtX1jObpfef02fvnn/UX3t3of157E/ZXv0EhjBAwAAAAAA9MOv/Hjd97+ZX6hqC/QrKwxi8oU7knTtLa/UyPibtHxhuaYFnLsFAQ8AAAAAAF2uUHiyeHqp4DmLp/L/7GWDA2W3Wyjc8fjf+CqNjL9Ji6eWCHlKIOABAAAAAKDLFZr+9Hs37yh4zvN51tp52eBA2bteXVxa1jfvP1ow3PEQ8pSHgAcAAAAAAOjmd9yU/X3v1l696b23FQxdJOnE489teu0P/vq2stq6uLQ27Wrx1JL+8GN7irYjEfKUg4AHAAAAAADo9ncF9NJdV0ha2yGr/5r8u2RJ0mz0nzZN0RoZf1PJoEbaHO5ce8vm7dbz8b/xVXrTe2/T4qklfeOjCUKeDdhFq0N95jOf0YULF1rdDWdt3bpV9913X6u7AQAAAABNtfOdL5dv4eX6cfxpffP+tS3LXzY4oP5r+nTFtl4tnlrSL586te6clw0OaGT8TWVPzfrGRxMVhzuem/b+viTpn//rD/WNjyaq2uWrUxHwdCjCndrw+QEAAADoVre849W65R2v1vNP/krPHPm5Fk8v6TepheyInd6tvXr5DQN62eCABv/tq8oatZPrir7eqsIdjxfyVLPDVycj4AHaUDKZ1NjYWENqB4NBTU9PN6R2PSWTSU1OTiqdTiuVSmlhYUEzMzMKBoOt7lrXisVimp2dVSqVyl4Xa22ruwUAAIAGufaWV1YdwhTzRx+v/e/0N+39/WzQgzUEPEAbSqVSSqVSkqRAIKD9+/fL7/fL5/NpYGBAPp9P6XRaiURCkUhEkuT3+zUzMyOfzydJ64KRubk5xWIxSdKxY8da86aq4Pf7lUwms59FMZFIRLFYTKFQSNFotO59aXR9FwwMDFR0TQAAAAA0DwEP0IYWFhYkSdFoVOPj4wWP8/v92YAnGAwqEAis+3nu95FIRENDQ0qn0w3ocf0FAoFs/wcHB0sGClNTU9lfGxHANLq+C0ZHRzU6OqqJiQlt37691d0BAAAAkINdtIA2lE6nFQgEioY7G3kjdwrx+/06ePBgNjxyid/vL3mMN3VrY8hVyNDQUEV9qLR+Jyt1rwEAAABoPgIeoA2dOXNG4XC47nVHR0edGcFTqdnZWc3NzWlubq6s4yudYlRpfQAAAABoJgIeoA2lUikNDw83pHYgEOjYkKeS0TXVfAaM3gEAAADQrliDB2hD6XS6rGlJ1fD7/VpYWOjqaTaJRKLVXQAAAADq5sEHH6zo+JWVFb344ot66Utfqi1btqz72Ysvvqif/exnNfWnWP16aGb997///XWv3yiM4AHaUCMDGL/f37EjeMrlwjbxAAAAQCN0UvjiYv1GIuAB2lCjpmdJ0u7du51caLleksmk4vF4q7sBAAAANJ3r4Yjr9RuNKVpAG2rkCJPR0dGG1W53yWRSe/bsaXU3AAAAgKZzPRxxvX4zEPAAXWZqakpPPPGEUqmU0um0fD5fdmeoWCym2dlZSWsLPUej0ez24BvFYrHseQsLC9mt3ScmJiqeXhaLxTQzMyNpbQqZd340Gi15biKR0PT0dPb9pFIpWWvXHROPxzU5OalkMrnudWPMpnrj4+Pr2i2nfj7pdDr7eXrvx/uMwuFwwTWWvHPyteddE2ntM0+lUvL7/ZqYmKhoAeh6XjsAAAC0P9fDEdfrNwsBD9Bl/H6/zpw5o2QymQ0IJCkSiWj37t2amZlRPB7X2NiYRkZGdPbs2U0P/WNjYwqHwwqFQutej0Qi2r59+6aQpJBEIqGxsTEFg0HNzMysayeVSmlsbEwTExNFawwMDMjv92ffTz6jo6PZkUupVEqDg4OSVFZQU079jWKxmCKRiKLRaDYw8yQSCQ0NDSkUCuX9jAq1F4/HNTs7u2l018jIiIaGhjQzM1PW6Kx6XTsAAAC4wfVwxPX6zUTAA3QZL+zYv3+/hoaGJK2FB7t3784GBMXW6BkbG1M0Gs07AiUajSqVSmlqair7fSFeiBQKhfJOSfP7/ZqZmdHY2FjRYCUQCGRHn2zfvr3gcZ6BgYGSx9RSPxKJaGpqSnNzc3lH1QSDQc3NzWloaEipVCo7csnjXZ9wOJwNopLJpJ544om8n9PMzIy2b9+uAwcOlAx46nXtAAAA4AbXwxHX6zcbiywDXSr3IX96enpdOBAKhTQ7O6v5+fl1o2rS6bTi8bjC4XDB0MULBrygIB9vdI7P5yu53tDBgwfLGjnT6KlF5dSPx+OamprS+Ph40SlTfr9f0Wg0e3yhYzyTk5MFAxefz6dAIKB0Ol10+/d6XTsAAAC4wfVwxPX6rUDAA3S5VCqlkZGRTa8Hg8FNIz2OHTsmaW2aUaGdqHLPKbQdezgclqSS06+ktQCj0Ho17ebAgQOSyntfoVBIPp9PkUik5Lb1u3fvLvpzb1RSsSCsXtcOAAAA7c/1cMT1+q1CwAOg4ELK+Y4LBALy+/1Fz/GCgnxTvVKpVHakSbnturDwbzweVzqdXrdIdCnDw8OS1kboFFNqAeXcRZwLqce1AwAAQPtzPRxxvX4rsQYPgIpGyHi7L5UjX+CQO3qkkp2f2t2hQ4ckVfZZescWm1pVac1iar12AAAAaG+uhyOu1281Ah4AVY+Q8dZ9mZ2d1cLCQnaqkDf6I98okPn5+ZrabFfeFuyVvC/v2I3bt29U6cLQ5ajm2gEAAKB9uR6OuF6/HRDwAF2umqAlmUwqEokokUgoFAopHA6vG42TSCQKjgDxgoNGhBatVM37uvrqqxvVnYJquXYAAABoT66HI67XbxcEPAAqMjU1pUgkIr/fr/n5+YqnD20cKdJuUqlUVVOiBgYGlE6nK3pfZ86cqbidWtR67QAAANCeXA5HXK/fTlhkGUDZvIDA5/Npbm6uqoBgcHBQUvuu8TI9PV3WtuwbeZ9FJe/LO7YZQUs9rh0AAADak6vhiOv12w0BD4CypNNpRSIRSVI0Gq14alcsFlMqlVq3g1M1QUq7Ghsbk3RpO/JyeMeOjo42pE+eel07AAAAtCcXwxHX67cjAh4AZcnd6Wnfvn1Fj80XBniLKwcCgeyaL6V2jypWr1HS6XRV6wPt27dPPp9P6XS67P56iyuHw+GK26tEva4dAAAAuoPr4Us3hjsSAQ+AMpW7tkxumFBoutLBgwclrY0mKSWZTNZ1OpfP58uOYMkXZqRSqaoWnvb5fNn3U877mpqakiSNj483fLpUPa8dAAAAOtvq6qrT4Uu3hjsSAQ/QMap9IC/3vNypVcVG3kxPT2t8fFzS+gAld/HiQCCgaDSqVCqVDToK9c1bFHhjvWJKvSdvFMvG95FOp8sKdwrVD4VCCoVCisViRT8jbyerYDBYVhhUq3peu40IggAAADrHysqKlpeXnQ1fujnckQh4AGelUqnsuiqSdPjw4bK3uE6n04rFYtnvY7FYyfP8fr9mZmYkSQcOHNgUFKRSKY2NjSkajWpiYkLSpQWLE4nEpoBgfHxc09PTikQiCofDm9qPx+M6cOCAZmZmsqFLNBpVLBZTPB7f1L/c1w4fPlz0vXjr0EQikXVBxoEDBwoGLuXWn56eVjQa1cjISN7wKhaLac+ePQqFQpqdnS1Yp9zr432+kjQ7O5v3uHpfu42fBSEPAACA+7xwpLe318nwpdvDHUky1tpW96EjDA8P20oWVy3lmWee0Y033lj1+Z/5zGd04cKFuvWn22zdulX33Xdfq7uxSTgcXvfgLykbfuR7yM7353toaCi79ku+c2dnZ9eN+NgolUppenpaiURCAwMD2Rq7d+9WKBRaN/3JC1D279+fHRmyUTqd1uTkZDZI8Na/GRoaUigUyv4+lUpl2xsYGMiGI2NjY9nAYeP7iUajRduNRCLZdn0+nyYmJrLrA3lqqe+9r9yt4YeHh9eNStpoZGQkG8BsbC8QCGhubk7SWsjiLezsHZt7HfNtg17rtSv2WQSDwaKBFdrD+fPntW3btoa2Uev/v4B8Hn30Ud15552t7gZQEe5bNNODDz5Y0/m54cjKyope+tKX1qlnm+u7Fu7cc889da1XD8aYOWvt8KbXCXjqo90CHlc042EDaATuXbiIgAeu4kEZLuK+RTPVEvBsDEe839eLy+GO5FbAwxQtAAAAAAC6kOvhSzPqu4SABwAAAACALtMJ4Usz6ruEgAcAAAAAgC7SKeFLM+q7hIAHAAAAAIAu0Unhi4v1G4mABwAAAACALuB6OOJ6/UYj4AEAAAAAoMO5Ho64Xr8ZCHgAAAAAAOhgrocjrtdvFgIeAAAAAAA6lOvhiOv1m4mABwAAAACADuR6OOJ6/WYj4AEAAAAAoMO4Ho64Xr8VCHgAAAAAAOggrocjrtdvFQIeAAAAAAA6hOvhiOv1W4mABwAAAACADuB6OOJ6/VYj4AEAAAAAwHGuhyOu128HBDwAAAAAADjM9XDE9frtgoAHAAAAAACHuRyOuF6/nRDwAAAAAADgMFfDEdfrtxsCHgAAAAAAHOZiOOJ6/XZEwAMAAAAAALJcD1+6MdyRCHgAAAAAAEDG6uqq0+FLt4Y7EgEPAAAAAADQWjiyvLzsbPjSzeGORMADAAAAAEDX88KR3t5eJ8OXbg93JAIeAAAAAAC6Wm440tNT/5iAcKc5CHgAAAAAAOhSrocvhDuXEPAAAAAAANCFXA9fmlHfJQQ8AAAAAAB0mU4IX5pR3yUEPAAAAAAAdJFOCV+aUd8lBDwAAAAAAHSJTgpfXKzfSAQ8aKlt27a1ugsAAAAA0BVcD0dcr99oBDwAAOcMDg4qkUi0uhsAAADOcD0ccb1+M1ze6g6gMd4+9V0tXFiu6JxvmT/TgM5WdM5/fM2I0pdXdvMfPP7XuupfK5vL+K3/N6KL/7pY0Tnnl98vu1pZOy/xXan/+JXRis5phnQ6rcnJSSUSCSWTSUmS3+9XMBjU2NiYgsFg9th4PC6fz7futWQyqbGxsYb0LRgManp6et1rxpiyz/f5fBoYGFAgENDIyIj27dsnn89Xl74lEgnNzMwolUpt+tnIyIhGR0fl9/vr0lYj26zk8/T4fD4NDw9rZGREoVCo4Gc6NjameDyePceTTqclSdPT0wqFQkX7le88STp79mzRa1no/Gg0qvHx8YLnJRIJpVIpRaPRdfd5pbZv3171uRuFQiFFo9GG1ZekEydObPo8c++NQtchEAhobm6urn0BAABucT0ccb1+sxDwdKhKwx1JFYc7kioOdyRVHO5IqjjckVRxuCNJ/yv924rPabRYLKZwOKzR0VFFo1ENDw/L5/MpnU4rkUgoGo1qenpaBw8elM/nUyQS2fSgmUqlsoFDIBDQ/v375ff7s+FKbr1IJCJpLUCamZnJPjSm02mlUiktLCxobm5OsVhMknTs2LFNfbbWKp1O69ixYxoZGZG09vA5Nze3Kdzw+jY7O6tIJKJwOJx9WK426InFYopEIhoeHlYkEtkUAqTTacViMQ0NDSkYDGY/u1o0ss1Cn+fRo0cVCAQ2He9dq0OHDmlyclKRSETj4+Ob7gtJ2TAqkUgoHA5nX5+eni4Ztp09e1YLCwuKRCLZkMi7b7z7q5izZ8/q8OHD2XZHR0c1MTGR9z1t7LO0FvSk0+mqr50Xemx877Ozs/L7/RoYGFh3/MLCgqS1ezaZTOrQoUPZwNX7tZz6c3Nz2T97+ep71292djb758xrd+Nnk+/ekNY+o2AwWLewFAAAuMv1cMT1+s1krLWt7kNHGB4etvkedKv1zDPP6MYbb6z6/NvvP1LxOT8wb634nD++5e0Vn3P4J39T8Tn/19N/VfE5i7/9UMXnSFL46/97Vec1wtTUlCKRSNFRFNLaqJ3JyUlNTExobGxMMzMzGh29NBLJC4lKjYxIp9PZUQehUGjTyJxcqVRKQ0NDGhgY0Pz8fMHjhoaGlEwmyx5FEIlENDU1VTTAKNb/PXv2KJlMbvoMCh0/NjamRCJR8jPe6Pz589q2bVtT25Qq/zxz2wsEAjp69GjBh/6RkZHstKdK/t+Qe9+Mjo5mA5hyhcPhbKBRju3bt5c1wqgSg4ODSqVS8vv9Re/njeLxuMbGxkqeV239dDqtoaEhpVKpkvdXOW14920j1fr/LyCfRx99VHfeeWeruwFUhPsWzfTggw/mfb2a8KKS3aNcD1/KqX/PPffUvd1aGWPmrLXDG19nDR6gTSWTSUUiEYVCoZIPsN7Ih0LTsNLptAKBQNFwZ6NS//Lv9/t18ODB7KiDeolGowqFQusebMvhBU7JZFJzc3MlgxZp7T3Ozs4qFAopHA5nRy+VqxVtVsrn82VHYiWTSe3Zs6fgsbn3jzcip9w2vBFL1a6Lk290UT7xeHzdlLNiIWQlvHqVjnjxRtaV+nNQbX1v5JukhrUBAAA6UzuEI91cvxUIeIA25T34l/vgOzo6WnA9kjNnzqybIlIvo6Oj69b6qJfch/Zy1g5Kp9MaGRnJjnKoZNSP114gENDU1NS6KTHt1ma1fD6fJiYmJK0Fh4XCm3379mV/f+jQoYra8N6/N9WvEseOHSv78zt06FB2Gp+09n4acQ9WYnx8vKF98Pl8CoVCFY38AQAA3c31cMT1+q1CwAO0qUQiUdY6JrkKhUGpVErDw5tG8NVFIBBoyMOtF1Ylk8m865vkGhsbUyqV0ujoaFmjaPLxphWFw+GS7UnSu971rqa3WYvc8G9ycjLvMbWMxMk9vpIpWslksux701ubxu/3a//+/dnXGx2QlSMYDJY92qwa3j0OAABQiuvhiOv1W4mAB2hD3oNcpVMtCo2CSKfTdd8tyuP3++s+Tcur6ym2vlU8Hs+GC+WOdirUnhfUHDhwoOix8Xhc3/3ud5vaZq1yP89iYZI3YqqSkThegOjdf4cPHy67X96InHLkLsgcCASy76le07RqEQgEGhrADA8PE/AAAICSXA9HXK/fagQ8QBvydtepZlRHvmlaCwsLDVuXw+/3t3SKjDeVLRgM1hxilTONqVVt1qrc6587TavckTjxeFzhcDg7qiadTpd973qLP5djZmZmXf+8sCd3l7hWGRwcbPg0LQAAgGJcD0dcr98OCHiANpT7MDc1NVXRufnWrGnU9CxJ2r17d0NG8OQ+sBfqfyKRyB6Xu0V0tXKDhkLTmFrRZj3kfp7FAhWfz1fxSJzZ2VkFAoF1i4GXM6qmkqmDXniS+2ej0vbqIR6P5w1yhoeH6/LnIHdE2kYbt1UHAADwuB6OuF6/XRDwAG3Km7oTiUQqGskTCoU2rQnTyIffYos718J7yB0dHS0YSORuq12vPuSu/ZPvQb4VbdZD7j2Uu35NPrkjcUpN0/LWxJEqD4e8LcbLEYvFNk3lyl0zqFnr8Bw6dChvkLMx4KpWKpUqeA9Uuv08AADoHi6HI67XbycEPECbyl3bZWhoSFNTUy3fLahZvAd5v99f9KE2d0pTpbtYFZJbJ1+40Yo268EbHeT3+zU+Pl702NyAsFSosDGkqWSa1uzsbNkh2aFDh/IuZu3dK5VMC6tFo/8MFtspi2laAACgEFfDEdfrtxsCHqBN+f3+dSNvIpGItm/fng17mvEw2wqRSESxWEyjo6Oam5sremwj1l25+uqri9ZvRZu18nbp8vv9JT9Tae3eK3ckzsaQJjeEKbbVeu7In1KKTeXKba/R07Sq2QK+UpUsUA0AAOBxMRxxvX47IuAB2lgoFNLMzMy6f7lPJpOKRCIaGhqSMUZjY2MNXZy30byRF16AlUwmNTs7u+l95zvPU88dwnJrbRxNkdvmdddd15Q2a5FIJDQ0NKRYLKZQKKT5+fmyR4GUMxInnU5vquf3+7Pvp9h9Wcn0rOnp6aI7bXkhT6PCkVQqlf0sa5VvBFBu/W4ZpQcAANqb6+FLN4Y7knR5qzsAoLjR0VGNjo4qkUhoZmZGhw8fXvcQGI/Hsw/SMzMzeaextINUKpX3gd4bFeHz+XTw4MGy+9+IhZ1LtdGKNgtJJpMaHBwsWmd4eFj79+/X0aNHK57eMzo6mt0t7NChQ3mnox0+fDjvej6jo6OamprK7m6VL4CbnZ0tOVXMk0gkim5HHw6Hs4sfJxKJmtZGSiaTMsZUfX4pqVSqofUBAABqtbq66nT40q3hjsQIHsAZwWBQ09PTOnv2rObn5xWNRjc9yI6NjTVtsdlKeevpbPyanZ2VtVbRaFQHDhzQ4OBgWdPPcncUqmfwkltr465Fud/Xc6RFsTYL8fv9mp+fL/h19uzZbIhSzdot5YzEKRQo5o62yTdtKt/In0KSyWTJwCYYDGbr1TpNy+/3y1q77mt+fl4zMzN1GSkWCAQ21T979qzm5uYaslg5AABAJVZWVrS8vOxs+NLN4Y5EwAM4yVso1wtHpqensw+43norrgmFQpqbm9PCwoKGhoZKrnWSGxDUM2zJnSK1cYRMK9psJS+8SaVSm+6pYiFNqXCo0MiffEpNz/J4O1jVOl0x33vy+/1lrQlVS5uBQECzs7N1nW4IAABQCS8c6e3tdTJ86fZwRyLgATqCF454D8DUoGEAACAASURBVKfe1BrX+P3+7FScctZnyX0YrlfgUmptn1a02Sq5wcrGBZNLhTS54dDGhaMrmUp4+PBhDQ4OyhhT9Gtqaip7TqPWpMrdlr1R2nWKJQAA6Gy54UhPT/1jAsKd5iDgAdpUpeGB3+/XwYMHJTVuq+1m2Ldvn6S1919qulnuw3a93nNunXwP861os1WKjcQpFdLkhj+551YyPSsej2tiYmLTlKZCX15fG7mbVqO3Ki81gis3yKqnRCLRttM7AQBAY7kevhDuXNLRAY8xxm+MCRljKvoncWNM0BjT2L/FA0WkUilNTk5WfN7o6Gj2IdfFaVrS+gfo2dnZosfmjvIpdWw50ul0drRJIBDI+zDfijZbqdBInFL9zH0vuaN/KpmedejQoezUq0r6mkgkGrYb1cjISNnrJFUjGAzmXdDaU2zr+VqkUqmGvi8AANCeXA9fmlHfJR0b8BhjopKCkg5LChpjynpKMMYEJdX+1AbUaOO0lnJ5D4fVnt9OSoVUwWAwG2jVY4vs3BoTExNt02Yr5RuJE4/HNTIyUvJcbzRWMpnMBi6zs7NlT0OqZLSPtH5KWaO2TA+FQg0N4XJHTeXTqBBmfn6+raYHAgCAxuuE8KUZ9V3SkQGPMWZcks9aG7PWpq21MUmDxpii4/Zzwp2ItbYx//wLlKna6T/eQ1q7jQSphNf3ckIqb82edDpd89orXi1vUd12arNVAoFA9p7yRo8cOnQoG94Ukxu4VDr9JxaLlbUOU67ccKSR07RaJZlMNiyESSaTTv83AwAAVKZTwpdm1HdJWwc8mSlWc8aY8dxpVjlTr2YLTL+akLTub/fW2oikAWNMNHNudgy8MSaQGfETlpSy1jZmkQOgAul0uqqQxxv1Mjw8XO8uNU0l25GPjo5m162pZlqbJx6PZwOlUuHA6Oio7rrrrqa22Upe8JRMJpVKpcoeWbNxmlY8Hi97etbMzExZIVKpvnaSRCLRsF3WEokEI3gAAOgSnRS+uFi/kdo64JHkkxSQFJU0b4yxxhgraV5SRFLYWrvub/DGmFGtjd7JN7cjlQl6DkvyZ4Ij75/MJzNtVfZPxkADVbMb1rFjx9pyLZdK5PY930N6vl2ZfD6fkslkVYvQptNpHThwQNLaiJpyFjr+yle+0vQ2WyU3lBkbG6toZE3uNK3p6emy3qcX6lVzD+eOGmrUblqtMj09XXR9nmqxuDIAAN3D9XDE9fqN1u4BjyTFJcUkJSQlM9+HrbWDG8OdDL+kov/kn5m2FbfWTmV+TWotREoUCIaAlkgmk+seWEuJRCJKp9PZaT+uyh1JkC/g2fiZ+Hy+7DbxkUik4gf7PXv2KJ1Oa3x8XOPj42Wd04o2WyU3MEwmkxWNrNkYBpUT2sRisbLW+MmnU6dpxWIxpVKpuo+ySafTikQijN4BAKALuB6OuF6/GVwIeGattWFr7Yi1dshaO5ZZU6eQQUmFxuXP59sdKzNda5+1tvwnaaAJvAf/sbGxklOVYrGYpqamND4+XpfRIPXahcgLaCqZLpMb4JS7U5Xf79eJEycUCAQ0NjZW1qiEdDqtoaGh7OiSSoOxVrTpWVhYqOq8anmhTqWjw4LBYPb4ckb+eIFDLXJ3/ip3Nznvfm/U7lu11I/H49k/E8WCmErbSKVSGhoaqngxawAA4B7XwxHX6zeLCwFPpdJam9qVj6/A4skzkg40rktAdSYmJjQ9Pa3du3fr+uuvVzgcVjwez+5KlE6nlUwmNTY2pnA4rGg0WnVgkEql1j1YHz58uOrtpsPhsMbGxjQ4OLjuoXNwcDDb12KCwWB29IU3csETiUQKBgDeqJrp6WlFIhENDQ3lXcconU5rampK27dv18DAgObn5yvajrtVbSYSiWxgUe0aTdXywply19DJ5QWOxUb+eAtWDw0NSVpb26ia95dKpdadd+DAgXW7eOWTSCTWBZHxeLyuQc/G+qVCJ+/PdSwW09DQUPazLxbubGyjUKB64sQJJRKJ7J9P7zhG8AAA0LlcD0dcr99Ml7e6Aw0wr7VpWvlcvfGFzI5bSWttZy3WAOeNj49n/1Xdm8YTi8V06NChdcGLt/tSNdsch8PhTaNOvDbT6fSmaTLW2rLqeud5YYDP56v4gTkUCml4eFiTk5MaGxvLvrf9+/eXHKEUCoUUCoWUSCQKjpIZGRmp69bQjWzTGLPue+/zzL0+5V6bannbw1cThIXD4aKjRNLptLZv3y5p7b3lvr9oNFr2FLbt27evu8+8NZK80Gh8fHzddcn3uUrrRxoFAgHNzc2V1f5GG+t7vP5UKt99U6iNShdjbtT26wAAoLVcD0dcr99snRjwHJY0bYwpNFonK7MD14Sk65vSM6BMfr8/b0DghQj1Mj093ZB1Suq13XcgENDMzEzV5weDwaYvXtyINhsd3pRrfn6+qvNKfSY+n68u7/Hs2bMVHd/oz7UZ163SNs6fP69t27Y1qDcAAKCduB6OuF6/FZyYopXZFn08s7151BgzY4zJ+7SQCXViWgtucmuMSjq04fAZSZFSQZCLBrb2VnzOgrZXfI7vX1cqPufc5S9WfM4Vl/dXfI7pqbydl/iurPgcAAAAAGgnrocjrtdvFRdG8IxIWrDWZvchziyUfNQYM51vwWVrbTgTAo1aa+OZcGcgdxqWMSaUObYj94f91vhdVZxV+aKt/6OKVvSnl35b7r8mv0M/rKYlAAAAAOgqrocjrtdvJdMuQ//zyUyhCuYLYTKhzYykoUJbm2dG+QS0YfvzTEB0InNuasPrQUlpa23J1T0zIVFIkq655pqhf/iHf6jk7RV11VVX6YYbbqhbvXb1u9/9TpdddlmruwFUjHsXLmrGffvss8/q3LlzDW0D3WdpaUl9fX2t7gZQEe5bNNOPfvQjSdLq6qqWl5fV29urnp7qJuysrq4WPLce9Uu13W71b7311rr3o1Z33XXXnLV2eOPrbR3wlGKMOSvpmLV2pOTB68+bkZSy1kZyXgtIikoKa20Xrv25Py9leHjYHjt2rJJuFPXMM8/oxhtvrFu9dsV6EHAV9y5c1Iz7tlv+/4XmevTRR3XnnXe2uhtARbhv0UwPPvhg3UameDU2cn1kTbX177nnnrr3pVbGmLwBjwtTtIpJSQqWs6CyJzPyJ2CtHct5zSfpqKTrvTrGmJQxJlpJyAMAAAAAQLO1azjSLfXbhROLLBfhTa+qZM/hg5LGNrw2obVpXNmQKPP7M5nwBwAAAACAtuRyOOJ6/XbS1gFPZg2eYryAZ9PQpAL1piUdzrNmz6ikJ/KcEpe0r5zaAAAAAAC0gqvhiOv1203bBjyZMGY+M6WqEG90TcntnzJr7Oyz1obz/NivS2FRVmYB5sEyugsAAAAAQEu4GI64Xr8dtW3AI2lAUjrzVewYScq7i9YGM5IObHwxZwpWWWv4AAAAAADQyVwPX7ox3JHaO+B5QmvbmBfbrjwoKZm71Xk+xpio1nbNim/8Wc66O6y1AwAAAADoaqurq06HL90a7kjtHfDEJBXcwcoYE9JaKLNxweSNx/klhUocl1aehZoz586X01kAAAAAAFy2srKi5eVlZ8OXbg53pDYOeDIja2aMMTMbF1vOrMsTlTRWavSO1qZmRUpsox6TtDvP66OSDlfQbQAAAAAAnOOFI729vU6GL90e7kjS5a3uQDHW2oQx5pikqDFmQJemUaUkXV8itPFG+chaGyvR1KSkE8YYn1czszbP1aXaAAAAAADAZbnhyMrKSkPrE+40TlsHPFJ2JE++na+KygQ0UUlD5bRhjNmjtRFDXlsRFZkiBgAAAACA6zaGI/UOeAh3mqftA54alTOFS5JkrU0aY8a0tnBzusB26k1lrZUxptXdAACgLNbaVncBAABUwPXwpRn1XdKxAU9m5E+xHbgKnbNpp61W6Onp0erqqi677LJWdwUAgLKsrq6qp6dtl/cDAAA5OiF8aUZ9l/C3sDa1ZcsWLS8vt7obAACUbXl5ueuHRgMA4IJOCV+aUd8lBDxtqq+vT+fPn291NwAAKNv58+fV19fX6m4AAIAiOil8cbF+IxHwtKn+/n4tLi6yngEAwAnWWi0uLqq/v7/VXQEAAAW4Ho64Xr/RCHja1BVXXKGenh4tLCy0uisAAJS0sLCgnp4eXXHFFa3uCgAAyMP1cMT1+s1AwNOmjDHauXOn0um0zpw5w0geAEBbstbqzJkzSqfT2rlzJ7s/AgDQhlwPR1yv3ywdu4tWJ9iyZYt27dqlkydP6uzZs+rv79e2bdvU29urnp4e/hINAGg6a61WV1e1vLys8+fPa3FxUT09Pdq1a5fTfyECAKBTuR6OuF6/mQh42tyWLVt0/fXX6+LFi1pcXNSvfvUrraysaHV1tdVdq4vf/va3uvLKK1vdDaBi3LtwUb3u256eHm3ZskV9fX269tprdcUVV/CPDgAAtCHXwxHX6zcbAY8DjDG68sordeWVV+oVr3hFq7tTV48++qhe//rXt7obQMW4d+Ei7lsAALqH6+GI6/VbgTV4AAAAAADoIK6HI67XbxUCHgAAAAAAOoTr4Yjr9VuJgAcAAAAAgA7gejjiev1WI+ABAAAAAMBxrocjrtdvBwQ8AAAAAAA4zPVwxPX67YKABwAAAAAAh7kcjrhev50Q8AAAAAAA4DBXwxHX67cbAh4AAAAAABzmYjjiev12RMADAAAAAACyXA9fujHckQh4AAAAAABAxurqqtPhS7eGOxIBDwAAAAAA0Fo4sry87Gz40s3hjkTAAwAAAABA1/PCkd7eXifDl24PdyQCHgAAAAAAulpuONLTU/+YgHCnOQh4AAAAAADoUq6HL4Q7lxDwAAAAAADQhVwPX5pR3yUEPAAAAAAAdJlOCF+aUd8lBDwAAAAAAHSRTglfmlHfJQQ8AAAAAAB0iU4KX1ys30gEPAAAAAAAdAHXwxHX6zcaAQ8AAAAAAB3O9XDE9frNQMADAAAAAEAHcz0ccb1+sxDwAAAAAADQoVwPR1yv30wEPAAAAAAAdCDXwxHX6zcbAQ8AAAAAAB3G9XDE9fqtQMADAAAAAEAHcT0ccb1+qxDwAAAAAADQIVwPR1yv30oEPAAAAAAAdADXwxHX67daTQGPMeZ3xph31KszAAAAAACgcq6HI67Xbwe1juAxkq6vR0cAAAAAAEDlXA9HXK/fLuoxRWvKGPMpY0x/HWoBAAAAAIAKuByOuF6/ndQj4PmxpH2Szhpj/qsx5lV1qAkAAAAAAMrgajjiev12U2vAk5b0KWvtDZL2SrpVUsoY8w8EPQAAAAAANJ6L4Yjr9dtRTQGPtXbAWvtQ5vcJa+2wpN2SrtZa0HPEGHNzHfoJAAAAAACawPXwpRvDHakB26Rba5PW2hGtBT3nJCWNMT8yxtxV77YAAAAAAED9rK6uOh2+dGu4IzUg4PFYa5OSntBayDMkKWGM+f+MMf++UW0CAAAAAIDqrKysaHl52dnwpZvDHakBAY8xpt8YM2mMOSPp05Ku8n4k6QZJXzPGnDHG3F3vtgEAAAAAQOW8cKS3t9fJ8KXbwx2pxoAns8bOLZnf9xtjJiWdlTQuabt3WOYrJmkw83pU0sHM1K1ttfQBAAAAAABULzcc6emp/0Qfwp3mqPXKjUgKGmO+oEvBjsn5uZE0JWm7tfY91toT1tpz1topSQOSfiEpWWMfAAAAAABAFVwPXwh3LqlHNBeVFNKlkTrS2ro7EWttj7X2g9bacxtPstamrbVjkp40xnyqDv0AAAAAAABlcj18aUZ9l9Rr7JUX7KQkhTPbp/9dmeeOSxqrUz8AAAAAAEAJnRC+NKO+S+oR8BitTbMas9beYK09WNZJxvy9MebnkvZI8tehHwAAAAAAoIROCV+aUd8ll9ehxri19jOVnGCMuV5r07qspGlJiTr0AwAAAAAAFNFJ4YuL9RupHiN4Kl4k2Vp7Qmvr9JjMr+E69AMAAAAAABTgejjiev1Gq3UEz4ikY1Wee52kYUnH8i3CDAAAAAAA6sP1cMT1+s1QU8BjrT1a7OfGmOskLVhrF/Oce05S0fMBAAAAAEBtXA9HXK/fLDVN0TLGfCET4uT72R6t7ap11hjzO2PMB2ppCwAAAAAAVMb1cMT1+s1U6xo8YRXYASszume7pAFJeyX9FSEPAAAAAADN4Xo44nr9ZqvHIssFWWvPZb4Skt4i6T2NbA8AAAAAALgfjrhevxUaGvBscJUKjPYBAAAAAAD14Xo44nr9VilrkeXMOjuFwpkRY0yx032ZcyckpSvoGwAAAAAAqIDr4Yjr9Vup3F20PigpJMnm+dl45qsc8TKPAwAAAAAAFXA9HHG9fquVNUXLWvserS2YvFfSQ5KKDtnZwOQcP1lR7wAAAAAAQEmuhyOu128HZa/B4y2WbK0dkzSstdDG6lKAU+grrbWRO8PW2ifr230AAAAAALqb6+GI6/XbRblTtNax1iaNMW+RdERS0Fr7SH27BQAAAAAAyuFyOOJ6/XZS9S5ama3PT9SxLwAAAAAAoEKuhiOu1283tW6THpZ0rB4dAQAAAAAAlXMxHHG9fjuqKeCx1h611i6Wc6wxZo8x5kwt7QEAAAAAgMZyPXzpxnBHqn0ET6V8TW4PAAAAAACUaXV11enwpVvDHanKRZZzGWOukxTR2s5axQKcgVrbAgAAAAAAjbGysqLl5WVt27bNyfClm8MdqcaAxxhzvaRnyz1ca9uqAwAAAACANuKFI729vU6GL90e7ki1T9GKai24KecLAAAAAAC0mdxwpKen/iu5EO40R61TtAKS0pJikuZLHDso6W9qbA8AAAAAANTJxnBkZWWlofXrjXDnkloDHr8kv7X2X0odaIzZI2m8xvYAAAAAAEAduB6+NKO+S2ode5WWtFDmsSmtLcYMAAAAAABaqBPCl2bUd0mtAU9Ca7tnlWStPWGt/bsa2wMAAAAAADXolPClGfVdUmvA82mtLbRckjHmemPMF2psDwAAAAAAVKmTwhcX6zdSTWvwWGuTxpiDxpgjkkatteeLHO6XFJL0V7W0CQAAAAD1cPzrX9Txrx/UrtvfqltDH6u6znM/OKLnHn9YSy+c1NILz2vlwqIu63+ZHvne72vHa+/QTX/y7jr2er0fxe7XQupn2vWGt+mG4D71bu0veOxC6mk99/2HdfKH39Hud/9n7XjdHQ3rF9qT6+GI6/UbraaAxxjzZknPSvqxpLQxJinpmNbW5tlosJa2AAAAAKBWp37yuE799HE9e3RGKxcWJUkLqZ9VXeuJL/5nLZ0+KUm65rVv0C3B/ZKkpx+b1emffk+nf/p9PfXVB3Tzn3+gIUHP0gsndfbEcZ09cVxPffUB9V2zU1tfca36XrFTvVu3aen08+uCJ8/A4Gvq3he0N9fDEdfrN0Otu2glJNnM743Wtk0PFDjW5BwLAAAAAA31bOKwnvvBt3Xhhed1cencuoCjVqd+8ri++8m7s9+/8f2f167b92a/f/7yV+ht9z2gIxOjWjp9Uk999QEtnT5Z00ihciydPqml0yd1Wt8veMxdH/pS0ZE+6DyuhyOu12+WeuyiZTJfyvl9vi8AAAAAaJqF1NM6/dPv6+LSOV3Rd5V23rZXuw98TNuvv6mmukunT64Ld27+8w+sC3c8vVv7ddeHvpT9fv7oYR3/+hdrarsWW7b2a+9knKlZXcb1cMT1+s1U6wieBUk+rY3kyTctK5df0utrbA8AAAAAynJr6GN5R8w894Nv11T3RwfvX/f9DcF9BY/tu2antl9/k86eOC5JeuqrD5RcK6cRdt62V7eGP87InS7jejjiev1mqzXgSUuattaWXDjZGBOUdKTG9gAAAACgZZZOn9Tpn16a/tR3zc6SocmO174hG/BI0vH/O6Zb/sN9de3X9utvWrfOzpat/ep7xbXa9Ya3adfte9V3zc66tof253o44nr9Vqg14Dkmaa7MY89IOlpjewAAAADQMs/9YP2/WW956baS5/Rds2vd9yd/+J26Bzxv/sh/a9jonO999n266U8PaMD/6qpreDuN/dt7P1/HnqEQ18MR1+u3Sk1r8Fhr32OtLWsSqbX2x9bat9TSHgAAAAC00nPff3jd932vKD0yprfvqnXfL50+qeU6LvjcaGf/5biOTIzq1E8er+r8535wRI997n06+y/HSx+Mmrkejrhev5VqXWS5bMaYq4wx72hWewAAAABQb7lTraTN4U0+vXlG+SzMV7c1eyvsnYyr75qd+u4n76445Dn1k8f12Ofep75rdmrvZLxBPYTH9XDE9fqt1rSAR9KwpJkmtgcAAAAAdZNv1E3v1tJTtPKFQEsvPF+XPuWzfGFRC6mndeonj2sh9XTNo4W83cC2bO3Xdz95txZST5d1nreVvBfusMBzY7kejrhevx2UXIPHGHOfpEFJUWvtv+T5WbluraxrAAAAANA+lpfO1a9WnadoLS+d07OJwzr+9YPZhZZz9V2zU4N79ummP3l3VfX7rtmpt07G9e2JUR2ZGNXeyXjRNXkId5rL9XDE9frtomjAY4w5IimY+XbMGOO31ub+12JKkm1U5wAAAACgEy0vpeta79sTo5Kkm/7kgHa89g3Z3b0WUk/rR7GP6uyJ43rqqw/o+NcP6s0f/nJVCyaXG/IspJ7Wdz95t7ZkRv4Q7jSey+GI6/XbSakpWiOZX42k7VqbZpUrnfPzcr4AAAAAwEn1HHWzfOF83WpJ0o7XvEGjX/6hbvqTd2vA/+psqDLgf7Xe/JH/pi2Z71cuLOrIxGjZ06w28kKeLVv79cjf/uWmOgupp3VkYlRbtvbrrZm1e9B4roYjrtdvN6UCnq/pUjCT1tq26LkWMr/GJcVKfCXq0F8AAAAAQMYNwf268Y/vLrr9eO/Wfu26/a3rXnvkb/+y6jb7rtmpN3/4y9k6XshDuNM6LoYjrtdvR0WnaFlrx4wx45L8WluDZ2NknZYUt9buL9WQMSYoaU/VPQUAAACANlPtSJxyFmcux67b92rX7XtLHrfjdXdo/ujh7PcrFxb1bOKwbgjuq6rdAf+r9eYPf1lHJkb1yN/+pW4NfVw/in2UcKdDuB6+dGO4I5Wxi5a1dspa+x5r7Yk8Pz4k6XCe1/M5o7URQQAAAADgnHquJdPb56tbrXL0veLaTa89mzhUU80B/6u1dzKulQuLeuxz75Mkwp0OsLq66nT40q3hjlTGLlrFWGv/roLD5yWN19IeAAAAALRKvu3Oy1HP3beqlS+cOnvieF1qb9nan3fnLrhnZWVFy8vL2rZtm5PhSzeHO1IZI3jqKCxppontAQAAAEDd5AtJqg1vBq6/qdbu6HuffZ/+z/036h//01t06iePV1WjloWjl06fzK7ls/vAx7RyYVHfnhjV0umTVddE63jhSG9vr5PhS7eHO1JzA55BSc0dhwgAAAAAdbRx+tHyi6UDkuUXN6/TU+s0pif/52d08odHJK0FLd/95N1V1al22tnS6ZP69sSoVi4s6s0f/rJuCO7TXR/6UjbkqeeOY2i83HCkp6f+MQHhTnPU5coZY95sjPmCMeZIga8nJB3QpW3VAQAAAMA5O297y7rvyxnBs3FEy5at/TUHPAt5plcV2/o8X+CypQ7hzt7JuAb8r5a0tpDzG9//+exW7IQ8bnA9fCHcuaSmNXgkyRhzQNLflzpMkpWUqrU9AAAAAGiVXW94m575f76U/X7phedLnrMx4Llhz1jRY49//YtaXjqnXXe8reAOWb0v3RzOeEFL3rp5+rnjNW8oeHyx/uULdzy7bt8rvf/zeuxz79ORiVHtnYzXdXFq1Jfr4Usz6rukphE8xpirJE173xb5kqRzkiZraQ8AAAAAWmnA/2ptz1k/Z+XCYsmRKgupn637/obg/rzHLZ0+qX/8T2/R/NHDOvnDI3rsc+/Tk//zM3mP3fG6O7K/3379TbrrQ1/Ke5wn3xo9N/3pgaLnbLR8YVHf/eTdBcMdz67b9+qN7/+8lk6fZCRPG+uE8KUZ9V1S6xSticyv5yTFJE1pbRrWVM5XQmsjd95srX2yxvYAAAAAoKVuDX183ffPJg4XPHbp9Ml1u1Xd+Md3F5yelW/b8tzRQrluCO5T3zU7tWVrv24NfXxd4LPR8oVFPfeDb697bXDPvqIjfvLVOJJZQPmuD32p5LmEPO2tU8KXZtR3Sa0BT1DSrLV2wFr7HmvtByUdk/Qpa+0HM19vkTQs6aAx5roa2wMAAACAlhrwv1pvfP/ns98/9dUH8oY8Gxc/3nnbXt3yH+4rWHf5wubFmIt54/s+p5ULi3rkb/9Sx7/+xQI1F/XIJ/5i3TbmO2/bq1tDHyu7nY3hTrEwKdeu2/dq94GPre229Ym/IORpE50UvrhYv5FqXYPHJ+ndG15Lay3QecR7wVqbNsZEJUUk/VWNbQIAAABASUunT+pHB+/f9PrGxYiXXng+u913rlsPfKzgaJtdt+9V74e+pCe++J+1dPqknjh4v577wbe147V3qHdrv849Nqt//ML3ssff+Md3Fw13vJrzR9cHRTtvy78Gj7QWNP27//Idfe9z79NTX31AT331AV3z2jeo7xU7tbx0TksvrB895I32KbSuTyGPfOIvKg53PDcE90mSnjh4vx75xF/orZ/+WkXno75cD0dcr99otQY81+eZdrUgaUQ5AY8kWWvjxhjW4AEAAADQFEunT+r0T7+f92e5O0itXFjMe1ypESc7XneH/t1/+Y6e+8ERPff4w1p64aSOf/2gVi4s6rL+l+ma175BO157h24I7itroWFvF6qnvvqALi6d047XvEG3hj9e9Jy+a3bqrZ/+mk795HE9mziss/9yPPtetmzt1/brb9KA/zXa8bo7Kg52PL19V1UV7ni8kGfjNDE0l+vhiOv1m6HWgMcYY/qttbn/5ZuTNK5L6/N4B14lyV9jewAAAABQlh2vu0N/duiZhrezG0IpiAAAIABJREFU6/a9m8KTRx99VHfeeWddapVjx+vuqDqAKeXNH/5yzTVuCO7LBj1oPtfDEdfrN0uta/CkJe0zxnzBGHNL5rWEpEFjzKc2HLsvczwAAAAAAGgC18MR1+s3U60jeI7p0jbpeyT9G2vtCWNMWlLEGONXJvDR2qieZI3tAQAAAACAMrgejrhev9lqHcEzJclkvq7Oef3TmdfGtBYAjWdeT9TYHgAAAAAAKMH1cMT1+q1QU8BjrU1I+jutjcz5YM7rU5JOaC3kUebXtCQWWQYAAAAAoIFcD0dcr98qtU7RkrU2UuBHAUlf1NrUrQVJ4Q2LMQMAAAAAgDpyPRxxvX4r1RzwFGKtPae1KVoAAAAA0PYefPDBkseU+3D44osv6mc/+1nFfXD94bZQ/XvuuafubWGzTr1/XKnfarWuwQMAAAAAXcH1h0/X66M416+v6/XbQUMDHmPMdcaY/ka2AQAAAACN5vrDp+v1UZzr19f1+u2ipoDHGPMFY8x1BX62R1JK0lljzO+MMR+opS0AAAAAaAXXHz5dr4/SXL6+rtdvJ7WO4AlL8uf7gbX2qKTtkgYk7ZX0V4Q8AAAAAFzi+sOn6/VRHlevr+v1201Dp2hZa89lvhKS3iLpPY1sDwAAAADqxfWHT9fro3wuXl/X67ejZi6yfJUKjPYBAAAAgHbi+sOn6/XRWq7fP916f5a1TXpmnZ1C4cyIMabY6b7MuROS0hX0DQAAAACazvWHT9fro7VWV1edvn+6+f4sK+CR9EFJIUk2z8/GM1/liJd5HAAAAAA0nesPn67XR2utrKxoeXlZ27Ztc/L+6fb7s6wpWtba92htweS9kh6SVHTIzgYm5/jJinoHAAAAAE3i+sOn6/XRWt717e3tdfL+4f6sYA0eb7Fka+2YpGGthTZWlwKcQl9prY3cGbbWPlnf7gMAAABAfbj88Ol6fbRW7vXt6an/Ur3cn81R7hStday1SWPMWyQdkRS01j5S324BAAAAQHO5+vDpen201sbru7Ky0tD69cb9eUnV0Vxm6/MTdewLAAAAALSMiw+frtdHa7l+/zSjvktqHXsVljRXj44AAAAAQCfphIdbwp3O5fr906z6LqlqipbHWnu0Xh0BAAAAgE7BVtNoZ67fP82s75KaV08yxtxnjDlkjLmlHh0CAAAAAJd5W013wsMt4U7ncf3+cb1+I9U0gscY82lJf5P59vWS/k3NPQIAAAAAR7HVNNqZ6/eP6/UbrdYRPCFd2g79XO3dAQAAAAA3sdU02pnr94/r9ZuhphE8GVOS/JIm61ALAAAAAJzDVtNoZ67fP67Xb5ZaY+Vjkn5krd1nrf1xsQONMa83xjxRY3sAAAAA0FZcf/hkq+nO1gn3j8v1m6nWETzvkfQdY8y8tfapEscOSArU2B4AAAAAtA3XHz7Zarqzdcr942r9Zqt1m/SUMWZE0t8bYyRpVlJS0kKew/21tAUAAAAA7cT1h0+2mu5snXT/uFi/FWrdRetMzrc+ScHaugMAAAAA7c/1h0/X66M416+v6/VbpdYpWkZrwY7d8FohtsjPAAAAAKDtuf7w6Xp9FOf69XW9fivVGvAsaC3gKXeL9KtqbA8AAAAAWsb1h0/X66M416+v6/VbrR7bpM9Ya/eXOsgYE5R0pA7tAQAAAEDTuf7w6Xp9FOf69XW9fjuodZv0VOarHFbFp28BAAAAQFty/eHT9foozvXr63r9dlHrLlpvqeDYo6o9UAIAAACApnL94dP1+ijN5evrev12UtfAxRhznTHmzcaYW3Jeu6XYOQAAAADQrlx/+HS9Psrj6vV1vX67qUvAY4x5tzHm55LmJc1KOpzz4//DGPNzY8yr6tEWAAAAADSD6w+frtdH+Vy8vq7Xb0c1BzzGmElJ05IGtbbGjvclSbLW7pP0iKRErW0BAAAAQDO4/vDpen20luv3T7fenzUFPMaYPZIiWgt0kpKmJIU3HmetDUs6Z4y5u5b2AAAAAKDRXH/4dL0+Wmt1ddXp+6eb789aR/CEJaUljVhrh621H7TWHixw7KclvafG9gAAQBMsLS8p9J279Wff3KdTF37V6u4AQNO4/vDpen201srKipaXl529f7r9/qw14PFLendmh6xSZjPHAwCANhd9YlKnXjylCysXdGHlQqu7AwBN4frDp+v10Vre9e3t7XXy/uH+rD3geb219qEKjvfV2B4AAGiw//70l/XUr59sdTcAoOlcfvh0vT5aK/f69vTUdbPtTfW5Pxun1itnjDHXlXmsX1KqxvYAAEADPfbL7+mhn3+t1d0AgJZw9eHT9fpoLdfvH+7PS2oNeFKS3lnmsWG1IOAxxviNMSFjTEXTw4wxQWMMI44AAF3j1IVfKfrEZKu7AQAt4+LDp+v10Vqu3z/NqO+SWgOeo5KmjDH/vthBxpgDkg5obaetpjHGRCUFJR2WFDTGhMo8L6i1NYMAAOgaH33sw9q6ZWuruwEAHaMTHm4JdzqX6/dPs+q7pNaA5++1tkV63BhzxhhzyBgzKclvjJnMfH8mc5wkNe2fBY0x45J81tqYtTZtrY1JGjTGTJc4zwt3ItbadDP6CgBAq33ksQ/p1IunFNk9UVOdT//oU5pPP1tTjcd++T39/+zdf4xd533n98+XlhglFsWhjPpHgmLFoZ0FHLSSSVotVmkhi5ScNPlnI46UPxI02LVEOXXXcGKTouH4j2wqmjSQwFvDtkjvbhc2FpZIp/0jbVcRGRDZyEjNH5K8stuNzaGabdaKA5FDWZK5HIlP/zjnipeXc889M8+995zPue8XMJjhnXM/51rP10cPH53zfD/37ceyMgCgabSaRpu51880851kLfCklJ6VdFjFIs+cpF2S9pR/3lP+eVN5+MGU0is55xsUEacrfr1P0jWLOSmlvZJujYgD5WNbW/uytpZ3/OyWtJhSOjjOzwoAQFt98/tH9fzfPae9H9yn9869LytrcemsPnHi43ruR8+u6f3P/M1f6MDJ/VpcOpv1OQCgSbSaRpu51497/iRlb4+dUtot6ZsqFnV6X6n8dZTfj6WU8v6T4IDyDp2tQ363S8XdOys9ErZYLvQ8qeJOoz3l8VJxh9FWSQvj/KwAALTVcz96Vv/qu/9Sv/a++3XXz/1idt4f3v0Fvftn3q3Pfuszq17kee5Hz+rAyf1698+8W3949xeyPwsANIFW02gz9/pxz5+0sfQ/SyktSLpP0lEVGyn3FnrOSNqdUvrwOM7TU26YXLVgNC+p8vGq8rGtoymlg+X3M5IOqFiMmupeQQAANKG3qfLt/9kd+q1f+Edjybx5/c36/bv+QG+/8e367Lc+U/txred+9Kw++63PvLW4c/P6m8fyeQBgmmg1jTZzrx/3/GkY21UnpXQspfRASum9KaV15df2lNLhcZ2jz25Jxyp+v0XDO3adXak7Vvm41gPlHUkAAHTeZ5/5jCRl77sz6N1vf4/+6O4v6O03vl2fOPHxkYs8LO4A6AL3v3x24S+3GM69ftzzp2X8y8oTVj5O9YSk8xWHLanYE2glc0M2Tz6iotMXAACd97lvP6aXXn9Jf3DXYxNZUKm7yHN26Qf67LeK7l2/f9cfsLgDwJL7Xz5pNd1tXagf5/xpyl7giYh7yo5Z9wz5/Z9GxFhaYZR33nywxiNUZ1U8prWSd6yQu0fSmZTS0cyPCABA6/2bF/9Pfes/PqPfvuNj2jL33omdp3+R5zPPfPq6RZ6zSz/QJ058XG+/8e36o7u/oHe//T0T+ywAMCnuf/mk1XS3daV+XPOnLWuBJyI2q3hUao+kPx1y2EFJD0TEl3LOVdqneq3Wnyw/37C7eN7St58Pd+8AADrv7NIP9KXnvqgP3/ZL+qXbfnni53v329+jP7ir+O88/Ys8LO4A6AL3v3zSarrbulQ/jvlNyL2DZ2/5PSRFRNwyeEC5N897Jd0XER9a64nKPXJODnm8avCcS5IOaWAj5r7Hu/odkbS3Ti4AAM4uXbmkzzzzaW3ZuEX/wx3/49TOu2XuvfqDux7Ta8uv6TPPfFrP/M1f6DPPfJrFHQDW3P/y6Z6Pau7j657flNwFnp0q7uBZknQopfRKxbGPll9rtXs1j1CVmyXP91qgl99v7X+8KyIeLo89lPG5AACwcPS1JyVJ//SusTw5vSpb5t6rP7r7C3pt+TUdOFncjMviDgBX7n/5dM9HNffxdc9v0g2Z75+XtDOl9GKNY09LenwtJykXYg6s9n0ppYWI2FnusXOsf4GofHzrgKRtA+eaU7FwtZRSqurU1ftcD0vSu971Lp04cWK1H3Hmvfrqq/xzgyVqF26efv0p/ejNv9X9Ny3o1LdODT3u0pVL17126tRp/Ycb/r/sz/C3b/ytfip+Sv8p/Se98cYb+su//L8097aRT1MDXHMxNXX2irly5YouX76s9evXa3l5eegGwleuXFnT3jN189eqqXz+PzxZvVobx/hW1W5X67OKU+3mLvBI1d2sBo9b9SyuXHCZSykNa3teqVykWWmh5rCKu47eyi0fAzugog37XEQcSCntXeG9vexDKh4F0/bt29Pdd9+9lo84006cOCH+ucERtQsnz/zNX+jfnfyO/pub/lv99/f+VuWxr15+VV/6P/7na17bvn1b9mbML732Qz1+4ku64YYb9I9/4SF96bkv6olL/5q7eFAL11xMywsvvFD5+95/+d+wYcPI//K/lr1nVpO/Fk3m8//hyXrhhRfGNr7DarfL9VnFqXZzF3iWJG2W9HyNY7dLWssizb6qRZa1KB/X2ppSWuh7bU7ScUmbe/vxRMTiqEUeAADa7uhfFY9m/dtLf65/+7/9+arf/4kTH7/mz//6v3tiVe3MX3rth/rEiY/rteXX9Ed3f0Fb5t6rd//Mu/XZb31GnzjxcR2+91/QHh1A67k/NuKej2ru4+ue3xa5CzyLKjYy/vUaxx7QKhd4yoWYp9fwuUY5LGnHwGv7VDzG9dZmyymlpYh4OSLm2IQZAODq9nfeoZvXb9CFCxe0adOmymP/9rWX9NLrL13z2paNW3Tz+g1rOvdKizuSdMc7P6C9H9ynAyf363dOfFx/ePcXWOQB0Fruf/l0z8dozuPrnt8muQs8xyV9MiLOSzq40l48EXGHigWVD6homb4aH5zA3TuPS3qyf7Pl0i6tvEfQUUkPqHwUCwAAN7/1C/9IUvmYy113Vx77ze8f1b/67r+85rWPfeCfrOkRrWGLOz13/dwvaq9Y5AHQbu5/+XTPRz2u4+ue3za5XbQeU9Eifbeks+XdLicj4qny+5sqNlfubWRce5PlcgPjXRFxeqUvFYsuGnhtVOZWSQ+UHbYGzWuFO4zKPXq21P3cAACg2Mvns898ZujiTs9dP/eL2vvBfXrp9Zf0Oyc+rlcvvzrlTwoAw7n/5dM9H/U5jq97fhtlLfCklC6qaH0e5UubJG1V0YVqa/l673cr3uFTkX0opbQlpbRtpS9Jp8rj+l8b5YikhwZfLPffkYo9hQAAQIZXL7+q3znxcb30+kv6/X/wByPv/mGRB0Abuf/l0z0fzXKvn1mtz9w7eJRSOijp87q6kDP4XZIOpJT25Z4rR0QckLTY3yq9p29/HXq1AgCQYXBx5453fqDW++76uV/Ub9/xMb30+kv6vWc+zSIPgEa5/+XTPR/N6rVJd62fWa7PcbRJV0ppb0R8RcWjWltVPO60pOIumwMppXPjOM9aRcS8pIdVdPwaZknF517pvWcn9NEAAGjU//Ldf6GzS1f/Nfe3r7103TFffPafXbPJ8t4P7hu6V87vPfPpVS/u9PzSbb8sSfrSc1/U7z3zaf3Rh/7Zqt4PAOPg/pdP93w0a3l5WZcvX554q3LqczLGssAjSeUizqPjyqvhVqlYgCn3yalyRNLeEZ2wDkn64Aqv7xIbLAMAOur5Hz2nsxev/+8Yb7/x7ZKk15ZfW/H3w9y8fsOaFnd6eos8z/zNX6zp/QCQw/0vn+75aFZvfNevX29ZP9TnGBd4pqFsm75PxV1CPWcjYlFFi/PrNk8uN2tWSmnUIs1+Sef6W6KXe/O8gxbpAICuGvddMv/0rv8pO+OXbvvltxZ6AGCanP/y6Z6PZvWP7/Ly8kTzqc/JmdoCT0TsUNGe/B1rzSj3z7luD52Kc85JOqCrXbyqspfKz3gkInoLRXvLLwAAAAAd5/qXT/d8NGtwfMe9wEN9Ts+07+BpYhPjhRqPcEmSUkpnImJBRRewpSHt1AEAAAB0kONfPt3z0Sz3+plGvpPsBZ6IuE3FXS7bVb2Ac2vuuVarfLTq2BreU/suIQAAAABYSRf+csviTne518+08p1kLfBExGZJP6h7uKSUcz4AAAAAcECrabSZe/1MM9/Jusz3H1CxcFPnCwAAAAA6r9dqugt/uWVxp3vc68c9f5JyH9HaKmlJRRvxUT1Ut0j6VOb5AACYKV/84hdXfH0tk4/XX39dL7zwQq1j3SdPdfI/9rGPjf28AECrabSZe/24509a7gLPvKT5lNKLow4sO1TtyTwfAAAzz31y454PAMPQahpt5l4/7vnTkPuI1pKk8zWPXRQtxwEAyOI+uXHPB4Bh3K9vXD+7zb1+3POnJXeB55iK7lkjpZTOpZQ+n3k+AABmlvvkxj0fAIZxv77RarrbulA/zvnTlLvA8zkVGy2PFBGbI+LLmecDAGAmuU9u3PMBYBj36xutprutK/Xjmj9tWQs8KaUzkg5HxFMRsWHE4fOSHs45HwAAs8h9cuOeDwDDuF/faDXdbV2qH8f8JmRtshwR90j6gaRnJS1FxBlJp1TszTNoS865AACYRe6TG/d8ABjG/frmno9q7uPrnt+U3C5axySl8udQ0TZ965Bjo+9YAABQg/Pkxj0fAIZxv76556Oa+/i65zdpHF20ovxS388rfQEAgFVyndy45wPAMO7XN/d8VHMfX/f8puXewXNe0pyKO3lWeiyr37ykD2SeDwCAmeI4uXHPB4Bh3K9v7vmo5j6+7vltkLvAsyTp8ZTSR0cdGBE7JT2VeT4AAJDBffI0C5MzAO3kfn1zz0c19/F1z2+L3Ee0Tkk6XfPYlyUdzzwfAABYoytXrlhPnmZlcgagfdyvb+75GM15fN3z2yTrDp6U0iOrOPZZSfflnA8AAKzN8vKyLl++rA0bNlhOnmZpcgagXdyvb+75qMd1fN3z2yb3Dh4AANByvcnN+vXrLSdPszY5A9Ae7tc393zU5zi+7vltNLUFnoh4KCJOTut8AADg2snNunXj/9c+kzMAXeV+fXPPR7Pc62dW63Oad/BsU9FJCwAATIH75GlWJ2cAmud+fXPPR7PYs89X1h48EfH9VRw+r9Gt1AEAwBi4T56mkQ8AK+nC9c05H81izz5vuW3S3yFp44hjou/nuczzAQCAEdwnT9PKB4BBXbm+ueajWezZ529cj2hFxZckpTGdBwAAVHCfPE0zHwAGdeX65piPZrFnXzfk3sFzXtLLkg4O+f2cpDsl7ZD0EUnnMs8HAACGcJ88uecD8Od6/XHPR7MGx3fcjzFTn9OTu8CzJOlkSulw1UERMS/pKRUbLQMAgDFznzy55wPoBsfrj3s+muVeP+zZd63ce6/2Szo06qCU0qKkfZIOZJ4PAAAM6MLkyTkfAIZxv75x/ew29/phz77rZS3wpJS+mVJ6tubhT0vamXM+AABwra5MnlzzAWAYWk2jzdzrhz37Vjb+3ZOGSCldVNEqHQAAjEGXJk+O+QAwTK/VtOv1jetnt7nXj3v+JE1tgSciPqBizx4AAJDJfXLjng8Aw9BqGm3mXj/u+ZOWtclyROyvcdhc+f0BFV23AABABvfJjXs+AAzTf/2ZxOasXD+Rw71+3POnIbeL1l5JaRXHj9yQGQAADOc+uXHPB4BhaDWNNnOvH/f8acl9RKv3yFXU+FpU0XULAACsgfvkxj0fAIZxv77RarrbulA/zvnTlHsHz3lJGyUdU/X+OiclHUopvZJ5PgAAZpL75MY9HwCGcb++0Wq627pSP67505a7wLMk6WhK6cFxfBgAAHA998mNez4ADON+faPVdLd1qX4c85uQu8BzqvwCAAAT4D65cc8HgGHcr2/u+ajmPr7u+U3JWuBJKT0yrg8CAACu5zy5cc8HgGHcr2/u+ajmPr7u+U3K3WQZAABMkOvkxj0fAIZxv76556Oa+/i65zeNBR4AAFrMcXLjng8Aw7hf39zzUc19fN3z24AFHgAAZoj75GkWJmcA2sn9+uaej2ru4+ue3xYs8AAAMCOuXLliPXmalckZgPZxv76552M05/F1z28TFngAAJgBy8vLunz5su3kaZYmZwDaxf365p6PelzH1z2/bVjgAQCg43qTm/Xr11tOnmZtcgagPdyvb+75qM9xfN3z24gFHgAAOqx/crNu3fj/tc/kDEBXuV/f3PPRLPf6mdX6ZIEHAICOcp88zerkDEDz3K9v7vloFnv2+WKBBwCADnKfPE0jHwBW0oXrm3M+msWefd5Y4AEAoGPcJ0/TygeAQV25vrnmo1ns2eePBR4AADrEffI0zXwAGNSV65tjPprFnn3dkDVyEXHbeD4GAADI5T55cs8H4M/1+uOej2a51w/1edWaF3gi4glJZyPi36ziPbdFxJcj4mREPBER/3Ct5wcAAFe5T57c8wF0g+P1xz0fzXKvH/bsu9YNGe9dKL+/WOfgiNgh6U/7XtomaVdEPJxS+ucZnwMAgJnWhcmTcz4ADON+feP62W3u9cOefdfLeUTrjKT5lNIjUq3HtY5IivLrnKRj5c+HIuKWjM8BAMDM6srkyTUfAIah1TTazL1+2LNvZTkLPE9I2l0+cvWmise13oyIbwweGBH3S5qTlCQ9nVJ6b0rpPknbVSzyLAy+BwAAVOvS5MkxHwCGodU02sy9ftzzJ2nNj2illD4fEVdULNpE368WImI+pXRn32sf7Pv5QF/GmYg4LOkBSTymBQBATe6TG/d8ABiGVtNoM/f6cc+ftJxNljf3fhz8laRtAxsoz/f9fGrg+K9IunWtnwMAgFnjPrlxzweAYWg1jTZzrx/3/GnIuerslrQkaWdKaV3vS8VjV+ckPdJ37Fzvh5TSKwM5i7p2AQgAAAzhPrlxzweAYdyvb1w/u829ftzzpyVngWerpD0ppT/rfzGldEbFI1e1Fm1SShfVtwAEAABW5j65cc8HgGHcr2+0mu62LtSPc/405Szw3KrrH7eS9NYiz2ruyrmY8TkAAOg898mNez4ADON+faPVdLd1pX5c86ctZ4Fn6KNV5f48/XvzDN1jpzz25YzPAQBAp7lPbtzzAWAY9+sbraa7rUv145jfhNwFngMR8ff6X4yIWyQ9Lunpmjm7yiwAADDAfXLjng8Aw7hf39zzUc19fN3zm5KzwPOEpC2SFiPi5Yj4fkR8X9IFSTskHZGkiNioYr+eVP75Q72A8nePqv5iEAAAM8V5cuOeDwDDuF/f3PNRzX183fObdMNa35hSejYivinpfhWbJPc2Su49mrU7Irap2HC59/qipKMR8bik85L2le87tNbPAQBAl7lObtzzAWAY9+ubez6quY+ve37T1rzAI0kppYWIOKJikadnUdK9KhZudqtomX5ExWNboWJj5r3lsSHp8RVapwMAAMlycuOeDwDDuF/f3PNRzX183fPbIGuBR3prkWejpO2SFlNK5/p+/cjg8RHxiKQDkjaqWNz5aO5nAAAA9bhPnmZhcgagndyvb+75qOY+vu75bZG9wCNJKaWLko7XPPaQeCQLAICpu3LlivXkaVYmZwDax/365p6P0ZzH1z2/TXI2WQYAACaWl5d1+fJl28nTLE3OALSL+/XNPR/1uI6ve37bTG2BJyJ2RMTL0zofAAAo9CY369evt5w8zdrkDEB7uF/f3PNRn+P4uue30Vge0VqFudGHAACAcemf3CwvL080n8kZmpYuLevy4dNKP3lD6x/apnWbfrrpjwRj7tc393w0y71+ZrU+sxd4IuI2FV2xtqt6AefW3HMBAID6Bic3417gYXKGtlk+8j2lC5eKP1x6o9kPA2vu1zf3fDSLPft8ZS3wRMRmST+oe7iklHM+AABQj/vkaRr56JblY2d15dyFpj8GOqAL1zfnfDSrt2ffhg0bLOtn1usz9w6eAyoWbgAAQEu4T56mlY/uePN7P9Kb3/oPTX8MdEBXrm+u+WgWe/b5y13g2SppSUXb87Mjjt0i6VOZ5wMAABXcJ0/TzEc3XLnwEy0f/V7THwMd0ZXrm2M+msWefd2Qu8AzL2k+pfTiqAMjYoekPZnnAwAAQ7hPntzz0Yzlrz8v3XQDe+5gLFyvP+75aBZ79nVHbpv0JUnnax67qGIzZgAAMGbukyf3fDTj8teeV7pwSTfuen9ezpHv6soPf5yV8eb3fqTLR76blYHmOV5/3PPRLPf6Yc++a+Uu8BxT0T1rpJTSuZTS5zPPBwAABnRh8uScj2a88cxf68q5C7px1/u17mc3ZGWll36sy4dP683Fuv/d8lpvfu9HWj76PaWX8haJ0D3u1zeun93mXj/s2Xe93AWez6nYaHmkiNgcEV/OPB8AAOjTlcmTaz6a8ebieb1xfFFv+wf/ud72/ndm561/aJti001a/vp3Vr3I8+bieS0f/Z5i001a/9C27M+C7qDVNNrMvX7Ys29lWQs8KaUzkg5HxFMRMeo/ncxLejjnfAAA4KouTZ4c89GM3qbK6zZv0o07t4wlM266UTf+xu3STTdo+evfqf241puL57X89e+8tbgTN1FnKPRaTbte37h+dpt7/bjnT1LWJssRcY+kH0h6VtJSRJyRdErF3jyDxvNvYAAAYD+5cc9Hc5a//rwk6caFvH13Bq3b9NNa/9A2XT58WpcPn9b6h7Zp3XuG//dLFncwDK2m0Wbu9eOeP2m5XbSOSUrlz6GibfrPJMeGAAAgAElEQVTWIcdG37EAAGCN3Cc37vlozuUj31W6cGliCyp1F3mu/PDHWv76d6SbbtCNv3E7izt4C62m0Wbu9eOePw3j6KIV5Zf6fl7pCwAAZHKf3LjnozlvnP6PuvJ//51u+JWfr7yzJldvkUc33aDLX3v+use1rvyw2JBZN91QLABt+umJfRZ4cb++cf3sNvf6cc+fltw7eM5LmlNxJ89Kj2X1m5f0gczzAQAws9wnN+75aM7P/Fh648//Sm/b+h7dsO1nJ36+dZt+Wut/83Zd/trzuvy157X+N2/XuvdsYHEHQ7lf32g13W1dqB/n/GnKXeBZkvR4Sumjow6MiJ2Snso8HwAAM8l9cuOej+akS8t6/7+T4j0368Zf/ftTO++692woFnkOn9blrz2vG3/157X8J3/F4g6u4359o9V0t3Wlflzzpy33Ea1Tkk7XPPZlScczzwcAwMxxn9y456NZl79WbKq8/jdvn/q5171nQ/G41qU3tHz0e8XnYHEHfdyvb7Sa7rYu1Y9jfhOy7uBJKT2yimOflXRfzvkAAJg17pMb93w0a/lP/r3SD1/VX/0X0tYmNzK+6Qbp0hvNnR+t5H59c89HNffxdc9vSu4dPLVFxMaI+LVpnQ8AgC5wnty456NZb37vR3rzzA91w455vbKpmc9w5cJP3rqD6IZf+Xnp0hu6fPi0rlz4STMfCK3hfn1zz0c19/F1z29S7h48q7Fd0hFJb5viOQEAsOY6uXHPR/PeeOavi+/HF/VfS7r05ydW9f7Lh6/dReCn9ty1qnbmVy78pMi49MZb7dJj001a/vp3dPnwaf3UP/mvaI8+o9yvb+75qOY+vu75TRu5wBMRn5S0RdKBlNKLK/yurjtX99EAAIDj5MY9H+2wbvMmpXIB5fyFC7p10/DbeNLST5QuXLrmtXjPzWtegFlpcUeS3jZ/q7Tr/Vo++j1dPnxa6x/axiLPjHG/vrnno5r7+Lrnt0HlAk9EPCVpZ/nHhYiYTym90nfIQUlpUh8OAACMl/vkaRYmZyjcuHPLWz//PydO6O67h2+y/MYzf603ji9e+/5f/ftvLcysxrDFnZ63vf+d0i6xyDOD3K9v7vmo5j6+7vltMWoPnnvL7yFpk4rHrPot9f2+zhcAAGjIlStXrCdPszI5Q3PSpWUtf/35oYs7PW97/zt14673K124pMuHTytdWp7yJ8W0uV/f3PMxmvP4uue3yahHtL4p6f7y5yUVbdH7nZe0UdLR8ucq85J2rPYDAgCAfMvLy7p8+bI2bNhgOXmapckZmpEuLReLNRcu6cbf+C9H3v3DnTyzw/365p6PelzH1z2/bSoXeFJKCxGxR8XizIGBx7OkYtHnaErpwVEnioidYoEHAICp601u1q9fbzl5mrXJGaZvcHHnbfO31nrf297/TqVfeUNv/O9/pctfe17rf/N2Fnk6xv365p6P+hzH1z2/jUZuspxSOljx6yckLVb8vt/LKu4IAgAAU9I/uVleHv9jJEzO0ITlY2eVfvjqW39OS9e3LV/+k39/zWLLjQvvH7r4cvlrz696cafnhm0/K0lvLfL81EODOxrAlfv1zT0fzXKvn1mtz6w26Smlz6/i2GclPZBzPgAAUN/g5GbcCzxMztCUK+cuXLPA85abyqntpTeUfvhq7U4gcdONumENizs9vUWeK9/7uzW9H+3jfn1zz0ez2LPPV9YCz2pExGZJD6eU9k3rnAAAzCr3ydM08uFr3HfJrP/N4R266rph289K5UIPvHXh+uacj2axZ5+3UV20xmle0p4png8AgJnkPnmaVj4ADOrK9c01H81izz5/lXfwRMQnx3iuO8eYBQAAVuA+eZpmPgAM6sr1zTEfzWLPvm4Y9YjWQan248sAAKBB7pMn93wA/lyvP+75aBZ79nXHqAWeJUkbJcWYzsdiEQAAE+A+eXLPB9ANjtcf93w0y71+2LPvWqMWeM6rWOA5Wv6cY17SjswMAAAwoAuTJ+d8ABjG/frG9bPb3OuHPfuuV+cOnqMppQdzTxQRO8UCDwAAY9WVyZNrPob74he/OLZ//q+//rpeeOGF6153r5+15n/sYx8b+2fB+NFqGm3mXj/s2beyUQs8xyT9YEznelnSN8eUBQDAzOvS5MkxH9Xcx9c9H82i1TTazL1+3PMnqXKBJ6X06LhOlFJ6VtID48oDAGCWuU9u3PMxmvP4uuejWbSaRpu51497/qSta/oDAACA1XGf3Ljnox7X8XXPR7P6x3fduvH/VYv6RA73+nHPn4ZRj2itSkTcIWm7pDkV+/ecSik9N85zAAAwy9wnN+75qM9xfN3z0SxaTaPN3OvHPX9axrLAUy7sHFHRKWvwd6clLaSU/t9xnAsAgFnlPrlxz0ez3OuH+uw29/qh1XS3daF+nPOnKfu+wYjYIem0isWdGPy1ijt6zkbE7bnnAgBgVrlPbtzz0Sy6EaHN3OuHVtPd1pX6cc2ftqwFnojYqOLOHenq4k70ffWf53DOuQAAmFXukxv3fDSr143ItX6oz25zrx9aTXdbl+rHMb8JuXfwPKxiv52QdFTSgqRtkjaV3xfK1yVpW0T8w8zzAQAwU9wnN+75aBbdiNBm7vXjno9q7uPrnt+U3D147pWUJG1bYTPlZ8uvb0bETkl/KunXJf2vmecEAGBmOE9u3PPRrP7xncT+HdQncrjXj3s+qrmPr3t+k3Lv4JmXtHtUp6yU0jFJj2qFTZgBAMBwrpMb93w0y71+qM9uc68f93xUcx9f9/ymjWOB51jNYw9J2pp5PgAAZorj5MY9H81yrx+6EXVbF+rHOR/V3MfXPb8Nchd4zqjYg6eOJOlc5vkAAEAG98nTLEzOZpl7/dCNqNu6Uj+u+ajmPr7u+W2Ru8BzStLOmsfuVLEgdI2I2BgRH8n8HAAAYARaTaPN3OuHbkTd1qX6cczHaM7j657fJrkLPAcl7a557KOSvrHC69slPZ75OQAAQAVaTaPN3OvHPR/V3MfXPR/1uI6ve37b5HbRukXSuYh4StWLNLtVbrAcEb828Ls7Mz8DAACoQKtptJl7/bjno5r7+Lrnoz7H8XXPb6PcBZ4zKvbWkeo9qnUk83wAAGAVaDWNNnOvH/d8VHMfX/d8NMu9fma1PnMf0Vrq+zkqvoYdo4HfAwCAMXGfPM3q5GxWuNePez6quY+vez6axZ59vnLv4DmvoovW0qgDR9iY+X4AANDHffJEq+lu60L9OOejmvv4uuejWb09+zZs2GBZP7Nen7kLPJJ0JKX04FrfHBE7JT01hs8BAADkP3mi1XS3daV+XPNRzX183fPRLPbs85f7iNZi+ZUjice0AAAYC/fJE62mu61L9eOYj9Gcx9c9H83qH99163KXCarzqc/JybqDJ6V0X+4HSCkdV/5CEwAAM8998uSej2ru4+uej3pcx9c9H80aHN9xP8ZMfU7P1BZWImJjRNwzrfMBADBL3CdP7vmo5j6+7vmoz3F83fPRLPf6Yc++a03zzpntkp6e4vkAAJgJXZg8Oeejmvv4uuejWe71Q312m3v9sGff9caxybIkKSJuU9FRa5it4zoXAAAodGXy5JqPau7j656PZtFqGm3mXj/s2bey7AWeiPiUpM+N4bMAAIBV6NLkyTEf1dzH1z0fzaLVNNrMvX7c8ycpa4EnIu6XdECjO2H1fp9yzgcAAArukxv3fIzmPL7u+WgWrabRZu71454/abl78Owrv49qc04bdAAAxsR9cuOej3pcx9c9H82i1TTazL1+3POnIfeqMy/piKQtkjallNYN+5KU3VIdAIBZ5z65cc9HfY7j656PZrnXD/XZbe71454/Lbl78MxJeiil9EqNY0c9xgUAACq4T27c89Es9/qhPrvNvX5oNd1tXagf5/xpyr2D54ykW+scmFI6LmlT5vkAAJhJ7pMb93w0i25EaDP3+qHVdLd1pX5c86ctd4HnmKRddQ9OKV3MPB8AADPHfXLjno9m9boRudYP9dlt7vVDq+lu61L9OOY3IXeBZ7+kRyLi7406MCJ2RMSbmecDAGCmuE9u3PPRLLoRoc3c68c9H9Xcx9c9vylZCzzlHTn3SToWEf84Im4Zz8cCAAASraabzEez6EaENnOvH/d8VHMfX/f8JuVusixJt0h6VtIhSYciYmkMmQAAQLSabiofzRoc33Fv0Ep9Iod7/bjno5r7+LrnNy1rgSciPiDpVO+P5feqjZRTzvkAAJg1jpMb93w0y71+6EbUbV2oH+d8VHMfX/f8Nsi9g2efaH0OAIAN98nTLEzOZpl7/dCNqNu6Uj+u+ajmPr7u+W2Ru8CzU9KSisezzko6X3HsnZI+mXk+AACwRrSaRpu51w/diLqtS/XjmI/RnMfXPb9Nchd45iTNp5ReHHVgRFyU9KnM8wEAgDXotZresGGD5eRpliZns8i9ftzzUc19fN3zUY/r+Lrnt01uS4JFVd+10++kpN2Z5wMAAKtEq2m0mXv9uOejmvv4uuejPsfxdc9vo9wFnjOSttc5MKV0MaV0OPN8AABgFWg1jTZzrx/3fFRzH1/3fDTLvX5mtT5zZ3qPSjpQ58CI2BwRX848HwAAqMl98jSrk7NZ4V4/7vmo5j6+7vloFnv2+cpa4EkpLUo6HBFPRcSGEYfPS3o453wAAKAe98kTraa7rQv145yPau7j656PZvX27HOtn1mvz6xNliPiHkk/kPSspKWIOCPplIrOWoO25JwLAADU4z55otV0t3WlflzzUc19fN3z0Sz27POX20XrmKRU/hyStpZfK4m+YwEAwAS4T55oNd1tXaofx3yM5jy+7vloVv/4TuIuV+pzOnL34FlSsXAT5Z+j4qsRETEfEQ9HxPwq37czIuYm9bkAABg398mTez6quY+vez7qcR1f93w0y71+qM+rcu/gOS9pTsWdPCs9ltVvXtIHMs+3KhFxQNJZSU9KeiAilFI6VON9OyU9LWnThD8iAABj4T55cs9HNffxdc9HfY7j656PZrnXD3v2XSt3gWdJ0uMppY+OOrBcNHkq83y1RcQeSXN9CzqHIuJARDyeUtpd8b7e4s7elNKoRSsAABrXhcmTcz6quY+vez6a5V4/1Ge3udcPe/ZdL3eB55Sk0zWPfVnS8bWeqFyw+WD5x96jUwdSSseGvGWfpB39L6SU9kbEkb47e06llM6U+VslPajiTqPFlNLBtX5WAACmpSuTJ9d8VHMfX/d8NItW02gz9/phz76VZS3wpJQeWcXhZ7WGNunlPjiHJe3vX3Tp3WkTEUdTSgsD79ml4u6dMytELpYLPXOSdpY5i+XXfhULVgsrvA8AgFbp0uTJMR/V3MfXPR/N6rWa3rBhg2X9UJ/d5l4/7vmTlHsHz2rslvSArt6FU9fhwQUcSUopHYuIg5L2RMSegTtu5jViT6Dy8auj/a9FxOOSjg1ZGAIAoDXcJzfu+RjNeXzd89EsWk2jzdzrxz1/0nK7aK3GFl19tKqW8i6bXRHx9JBDeq8/uMK5Foe85+xK3bHKR7QeqNqfBwCANnCf3Ljnox7X8XXPR7P6x3fduvH/VYv6RA73+nHPn4ax3METEfeoeKxpWCvyWyVtlbTaO2NuLb9vH/L78+X3wQWbpRVe65kbsnnyEUkPre7jAQAwXe6TG/d81Oc4vu75aNbg+I67+w71iRzu9eOePy3ZCzwR8ZCkr4w6TFLS8LtqVpRSWoyIbRr+uFVvQWlwo+WzGr7Y9I7rPlyxgfOZlNLRFY4HAKAV3Cc37vlolnv9UJ/d5l4/tJruti7Uj3P+NGXdNxgRGyU93vtjxZckXVSxifGqpJTOpJSGLQz1Hs16fOD1J8vPN/KRsIiYV9Fxi7t3AACt5T65cc9Hs+hGhDZzrx9aTXdbV+rHNX/acu/g2Vd+v6hiUWVJRaesQ33HbFVxN82ulNJzmed7S7lnzi5Jewc3RU4pLUXEofLz7e17zy5JTwxEHSkzKjdlBgCgKe6TG/d8NItuRGgz9/qh1XS3dal+HPObkLvAs1PS0ymlD/deKBdeHkspvdL32pyKluYLKaUXc07Ya2+uYvFmYdhjVSml3RFxJCJ2pZSOlos7t/YfHxEPl8ceWikDAICmuU9u3PPRLLoRoc3c68c9H9Xcx9c9vym5Czxzkj4y8NqSik2R/6z3QnlHzQEVd9N8dC0nKheOHizPOa/iTpzKTZtTSgsRsbPcY+fYwOLOnKQDkrYNnKe3gLSUUhrc2wcAgKlynty456NZ/eM7if07qE/kcK8f93xUcx9f9/wmRUpp7W+OeDOl9LaB174i6UJKad8Kx38/pfS+NZ/w2qytko5L2p9SOriG9x+RtJhS6n+Ea6uKRZ/dKhaSHuz//QoZD6t4JE3vete7tn3jG99Y7ceYea+++qpuvvnmpj8GsGrULqblL//yL8fWavfKlStvZV25ckWXL1/W+vXrJ9LK1yH/zjvvHPOnQr9vf/vba37v4Pj21+44ONRnFWp3curUbd3xXWvdutfnsHzqdrJ6tTuO8a2q3a7WZ5U21u6HPvSh0yml67qN597BExFxS//jWJJOS9qjq/vz9A7cqOGdrVYtpXSm7OB1JCLeUbUQM6h8XGtrSmmh77U5FQtGm3v78UTEYkQcGJZdPtp1SJK2b9+e7r777rX/D5pRJ06cEP/c4IjaxbS88MILY8vqvxvi9ddfn/ieJm3P5//Dk7XW2l1pfMe5h4dLfVblU7uTM6puVzO+a6nbLtTnsHzqdrJeeOGFsY3vsNrtcn1Wcard3CWxJUkPRMSXI+KO8rVjkrZExGMDxz6g4e3O16R85GpJ0p6yG1ZdhyUtDLy2T8VjXG99xvLnl+t04wIAwIH7bc9dvq0a/vVDN6Ju60r9uOajmvv4uue3Re4CzykVLcofVtmaPKV0TsWiy96I+EZEfCQi9kv6iqRh7c5zP4NUdNQaKSIel/TkYOet8v0nV3jLURWLUwAAWKPVNNrMvX7oRtRtXaofx3yM5jy+7vltkrvAc1BSlF/v6Hv9c+VrCyoWgPaUr69q0+KyC9aFEXfn9O64+WCNvK2SHkgp7V7h1/NaYQEqpbQoaUudzwsAQFv1Wk27Tp5maXI2i9zrxz0f1dzH1z0f9biOr3t+22Qt8JRdpj6vopvVo32vH5R0TsUij8rvS5L2r/IUu1Rsdryz4pje41Mr3X0z6IikhwZf7HsEa6yPkAEA0Aa0mkabudePez6quY+vez7qcxxf9/w2yt1kWRWbG2+V9FVJOySdl7R7YDPmOo5JOlJuZjxMb+foypbpZZv2xf5W6T1lG3fp6mIRAACdQKtptJl7/bjno5r7+Lrno1nu9TOr9Zm9wDNMSumirt/IeLUOSLp32C/Lblhzko6WdxMNO25exT5BmyvOtaQVunyV7z1b9wMDANAWg5ObcS/wMDlDDvf6cc9HNffxdc9Hs9izz9f4G8uPUbloc7bci+eaxZeI2KmiG9ax/nbnQxyRtLe/Q9YKDmnlfXx2qdxAGgAAF+6Tp2nkozldqB/nfFRzH1/3fDSLPfu8Zd3BU7ZGv7X84/mU0nPl67epWFTZWv7utKSHUkrPr/YcKaVDEfGkpAMR0TtX71Gqh1Z65GrgMz7cyxlxqv2SzkXEXG8hqNyb5x0jFoYAAGgV98kTraa7rSv145qPau7j656PZrFnn7/cR7SOqnjsKVQ8xvS+iNioYj+cjbq6yfJ2SccjYn4N+/CoXGBZqfNVpXKB5oCkbXXOERE7JB2JiN659pZfAABYcJ880Wq627pUP475GM15fN3z0Sz27OuG3Ee0Hpd0UdK9KaX3la/t09U7bI6klNapuMvnRfV12pqihbLV+UgppTMq9g3aKmk+pbSbu3cAAC7cJ0/u+ajmPr7u+ajHdXzd89Es9/qhPq/KXeDZKekjKaXjfa/t6fv5IemtO3AeVcWGyZOQUlqq2ny54j2VmzYDANA27pMn93xUcx9f93zU5zi+7vlolnv9sGfftXIXeLaraGUuSYqIXpeqpGLz4/7HsU7q6p48AABgTLoweXLORzX38XXPR7Pc64f67Db3+mHPvutld9EaWMTZ2ffzmYHjLuaeCwAAXKsrkyfXfFRzH1/3fDSLVtNoM/f6Yc++leUu8MTAn/s3Mz55zYHF3T0s8gAAMCZdmjw55qOa+/i656NZtJpGm7nXj3v+JOUu8CxGxO19fx56B4+KLljsawMAwBi4T27c8zGa8/i656NZtJpGm7nXj3v+pOW2ST8u6asRsVfFBsrzKvbfOZNSerF3UETcL+lTkrZkng8AgJnnPrlxz0c9ruPrno9m0WoabeZeP+7505B7B89jkt4r6Wld2z1rQZIiYkdEnJT0ZPn6/ZnnAwBgprlPbtzzUZ/j+Lrno1nu9UN9dpt7/bjnT0vWAk+5cfJWSX8s6ZyKR7C29929syBpU/m7c5J+Ped8AADMMvfJjXs+muVeP9Rnt7nXD62mu60L9eOcP025j2gppXRO5R07K/zukdx8AADgP7lxz0ez6EaENnOvH1pNd1tX6sc1f9qy26QDAIDJcp/cuOejWXQjQpu51w+tprutS/XjmN8EFngAAGgx98mNez6aRTcitJl7/bjno5r7+LrnN4UFHgAAWsx5cuOej2b1j++6deOfslKfyOFeP+75qOY+vu75TWKBBwCAFnOd3Ljno1nu9UN9dpt7/bjno5r7+LrnN40FHgAAWsxxcuOej2a51w/diLqtC/XjnI9q7uPrnt8GLPAAADBD3CdPszA5m2Xu9UM3om7rSv245qOa+/i657cFCzwAAMwIWk2jzdzrh25E3dal+nHMx2jO4+ue3yYs8AAAMANoNY02c68f93xUcx9f93zU4zq+7vltwwIPAAAdR6tptJl7/bjno5r7+Lrnoz7H8XXPbyMWeAAA6DBaTaPN3OvHPR/V3MfXPR/Ncq+fWa1PFngAAOgo98nTrE7OZoV7/bjno5r7+Lrno1ns2eeLBR4AADrIffJEq+lu60L9OOejmvv4uuejWezZ540FHgAAOsZ98kSr6W7rSv245qOa+/i656NZ7Nnnb2oLPBGxOSL2T+t8AADMIvfJE62mu61L9eOYj9Gcx9c9H81iz75umOYdPPOS9kzxfAAAzBT3yZN7Pqq5j697PupxHV/3fDTLvX6oz6tuqPplRHxyjOe6c4xZAACgj/vkyT0f1dzH1z0f9TmOr3s+muVeP+zZd63KBR5JByWlaXwQAACwNl2YPDnno5r7+Lrno1nu9UN9dpt7/bBn3/VGLfAsSdooKcZ0PhaLAAAYo65MnlzzUc19fN3z0SxaTaPN3OuHPftWNmqB57yKBZ6j5c855iXtyMwAAAClLk2eHPNRzX183fPRrF6r6Q0bNljWD/XZbe71454/SXXu4DmaUnow90QRsVMs8AAAMBbukxv3fIzmPL7u+WgWrabRZu71454/aaMWeI5J+sGYzvWypG+OKQsAgJnlPrlxz0c9ruPrno9m9Y/vJDZnpT6Rw71+3POnoXKBJ6X06BjPdVa0SQcAIIv75MY9H/U5jq97Ppo1OL7jXuChPpHDvX7c86dl3RTPtVvSkSmeDwCATnGf3Ljno1nu9UN9dpt7/dBqutu6UD/O+dM0zQWeLZLmpng+AAA6w31y456PZtGNCG3mXj+0mu62rtSPa/60jdqDp5aIuEfSgopOWSu5VdJWSWfGcT4AAGaJ++TGPR/NohsR2sy9fmg13W1dqh/H/CZkL/BExEOSvjLqMElJ0mLu+QAAmCXukxv3fDSLbkRoM/f6cc9HNffxdc9vStYjWhGxUdLjvT9WfEnSRUn7c84HAMCscZ7cuOejWf3ju27d+HcVoD6Rw71+3PNRzX183fOblHsHz77y+0VJT0pakvSwpEN9x2xV8ejWrpTSc5nnAwBgprhObtzz0Sy6EaHN3OvHPR/V3MfXPb9puQs8OyU9nVL6cO+FiNgq6bGU0it9r81JejoiFlJKL2aeEwCAmeE4uXHPR7Pc64duRN3Whfpxzkc19/F1z2+D3Ptd5yTtHXhtSdL2/hdSSkuSDqxwLAAAmCL3ydMsTM5mmXv90I2o27pSP675qOY+vu75bZG7wLN5hceuzku6d/DAlNJRFXf8AACABtBqGm3mXj90I+q2LtWPYz5Gcx5f9/w2yV3giYi4ZeC105J2rXDgRg1vow4AACao12radfI0S5OzWeReP+75qOY+vu75qMd1fN3z2yZ3gWdJ0gMR8eWIuKN87ZikLRHx2MCxD5THAwCAKaLVNNrMvX7c81HNfXzd81Gf4/i657dR7ibLp3S1TfoOST+fUjoXEUuS9kbEvMoFH0l7JJ3JPB8AAFiF/snNJDZnZXKGHO71456Pau7j656PZrnXz6zWZ+4dPAclRfn1jr7XP1e+tqBiAWhP+fqxzPMBAICa3CdPszo5mxXu9eOej2ru4+uej2axZ5+vrAWelNIxSZ9XcWfOo32vH5R0TsUij8rvS5L255wPAADU4z55otV0t3WhfpzzUc19fN3z0Sz27POW+4iWUkrDWp9vlfRVFY9unZe0O6X0Su75AABANffJE62mu60r9eOaj2ru4+uej2axZ5+/7AWeYVJKF1U8ogUAAKbEffJEq+lu61L9OOZjNOfxdc9Hs9izrxty9+ABAAAt4T55cs9HNffxdc9HPa7j656PZrnXD/V5VfYCT0TcExH7I+KeIb//0xVapgMAgDFynzy556Oa+/i656M+x/F1z0ez3OuHPfuulfWIVkRsVtEZK0n61JC8g5K+EhFzKaXfzjkfAAC4XhcmT875qOY+vu75aJZ7/VCf3eZeP+zZd73cO3h6GyyHpIiIWwYPSCkdSym9V9J9EfGhzPMBAIA+XZk8ueajmvv4uuejWbSaRpu51w979q0sd4Fnp4o7eJYkHRrRJetR9bVSB8yMDYEAACAASURBVAAAebo0eXLMRzX38XXPR7NoNY02c68f9/xJyu2iNS9pZ0rpxRrHnpb0eOb5AACA/Cc37vkYzXl83fPRLFpNo83c68c9f9LG0UXr/CqOmxvD+QAAmGnukxv3fNTjOr7u+WhW//iuWzf+hsXUJ3K41497/jTkXnWWJG2ueex2SYuZ5wMAYKa5T27c81Gf4/i656NZ7vVDfXabe/24509L7gLPoqR9NY89IBZ4AABYM/fJjXs+muVeP9Rnt7nXD62mu60L9eOcP025CzzHJS1ExJci4raVDoiIOyLipKQPSDqTeT4AAGaS++TGPR/NohsR2sy9fmg13W1dqR/X/GnLXeB5TEWL9N2SzkbEyxFxMiKeKr+/qWJz5W3l8WyyDADAKrlPbtzz0Sy6EaHN3OuHVtPd1qX6ccxvQtYCT0rpoorW51G+tEnSVhXt07eWr/d+d7Bmty0AAFByn9y456NZdCNCm7nXj3s+qrmPr3t+U7K3dk8pHZT0eV1dyBn8LkkHUkp19+oBAAAl58mNez6aRTcitJl7/bjno5r7+LrnN+mGcYSklPZGxFdUPKq1VdK8ig5bp1Qs7pwbx3kAAJg1rpMb93w0a3B8x71BK/WJHO71456Pau7j657ftLEs8EhSuYjz6LjyAAAAraabyEez3OuHbkTd1oX6cc5HNffxdc9vg/Hf7zpERGyMiF+b1vkAAMD13CdPszA5m2Xu9UM3om7rSv245qOa+/i657fF1BZ4JG2XdGSK5wMAAH1oNY02c68fuhF1W5fqxzEfozmPr3t+m4x8RCsiPilpi4q9dF5c4Xd13bm6jwYAAMal12p6w4YNlpOnWZqczSL3+nHPRzX38XXPRz2u4+ue3zaVCzwR8ZSKlueStBAR8ymlV/oOOSgpTerDAQCAfLSaRpu51497Pqq5j697PupzHF/3/DYa9YjWveX3kLRJxWNW/Zb6fl/nCwAATBGtptFm7vXjno9q7uPrno9mudfPrNbnqEe0vinp/vLnXtvzfuclbZR0tPy5yrykHav9gAAAYG1oNY02c68f93xUcx9f93w0iz37fFUu8KSUFiJij4rFmQMDj2dJxaLP0ZTSg6NOFBE7xQIPAABT4T55otV0t3WhfpzzUc19fN3z0Sz27PM2cpPllNLBil8/IWmx5rleVnFHEAAAmCD3yROtprutK/Xjmo9q7uPrno9msWefv5ELPFVSSp9fxbHPSnog53wAAKCa++SJVtPd1qX6cczHaM7j656PZvWP7yTucqU+p2P8uy0CAIBGuE+e3PNRzX183fNRj+v4uuejWe71Q31exQIPAAAd4D55cs9HNffxdc9HfY7j656PZrnXD3v2XWuiCzwRcVtE3DLJcwAAMOu6MHlyzkc19/F1z0ez3OuH+uw29/phz77rZS3wRMSXI+K2Ib/boWID5gsR8WZE/G7OuQAAwPW6MnlyzUc19/F1z0ezaDWNNnOvH/bsW1nuHTy7VbRQv05K6bikTZJulfRhSR9lkQcAgPHp0uTJMR/V3MfXPR/N6rWadq0f6rPb3OvHPX+SJvqIVkrpYvl1TNJ9kh6Z5PkAAJgV7pMb93yM5jy+7vloFq2m0Wbu9eOeP2nT3GR5o4bc7QMAAOpzn9y456Me1/F1z0ez+sd33brx/1WL+kQO9/pxz5+GG+ocVO6zM2xx5t6IqHr7XPnefZKWVvHZAADAAPfJjXs+6nMcX/d8NGtwfMfdfYf6RA73+nHPn5ZaCzySHpX0sKS0wu/2lF91HK15HAAAGOA+uXHPR7Pc64f67Db3+qHVdLd1oX6c86ep1n2DKaVHVGyY/GFJfyyp8padAdF3/P5VfToAACDJf3Ljno9m0Y0IbeZeP7Sa7rau1I9r/rTVfjC0t1lySmlB0nYVizZJVxdwhn0tqbhzZ3tK6bnxfnwAALrPfXLjno9m0Y0IbeZeP7Sa7rYu1Y9jfhPqPqJ1jZTSmYi4T9JTknamlP5svB8LAABI/pMb93w0i25EaDP3+nHPRzX38XXPb8qat3YvW5+fG+NnAQAAA5wnN+75aBbdiNBm7vXjno9q7uPrnt+k3H9b7pZ0ehwfBAAAXM91cuOej2a51w/12W3u9eOej2ru4+ue37Q1PaLVk1I6Pq4PAgAAruc4uXHPR7Pc64duRN3Whfpxzkc19/F1z2+DrDt4IuKOiLin/Lqj7/XbIuJkRLxZfn07Im7P/7gAACCH++RpFiZns8y9fuhG1G1dqR/XfFRzH1/3/LbIuoNHRXeszSq6ZZ2V9L6I2CjpjKSNutoefbuk4xExn1J6JfOcAABgDWg1jTZzrx+6EXVbl+rHMR+jOY+ve36b5O7B87iki5LuTSm9r3xtn6S58ucjKaV1km6V9KKkRzPPBwAA1oBW02gz9/pxz0c19/F1z0c9ruPrnt82uQs8OyV9ZGAvnj19Pz8kSSmlJRWLO/dmng8AAKwSrabRZu71456Pau7j656P+hzH1z2/jXIXeLZLOtb7Q0RsLn9Mko4NPI51UtLWzPMBAIBVoNU02sy9ftzzUc19fN3z0Sz3+pnV+sye6Q0s4uzs+/nMwHEXc88FAADqc588zerkbFa41497Pqq5j697PprFnn2+chd4YuDP2/p+PnnNgcXdPSzyAAAwBe6TJ1pNd1sX6sc5H9Xcx9c9H81izz5vuQs8iwPtz4fewSNpt/oe5wIAAJPhPnmi1XS3daV+XPNRzX183fPRLPbs85fbJv24pK9GxF4VGyjPq9h/50xK6cXeQRFxv6RPSdqSeT4AAFDBffJEq+lu61L9OOZjNOfxdc9Hs/rHdxJ3uVKf05F7B89jkt4r6Wld2z1rQZIiYkdEnJT0ZPn6/ZnnAwAAQ7hPntzzUc19fN3zUY/r+Lrno1nu9UN9XpW1wFNunLxV0h9LOqfiEaztfXfvLEjaVP7unKRfzzkfAABYmfvkyT0f1dzH1z0f9TmOr3s+muVeP+zZd63cR7SUUjqn8o6dFX73SG4+AACo1oXJk3M+qrmPr3s+muVeP9Rnt7nXD3v2XS+7TXpdEbExIn5tWucDAGAWdGXy5JqPau7j656PZtFqGm3mXj/s2beyqS3wSNou6cgUzwcAQKd1afLkmI9q7uPrno9m0WoabeZeP+75kzTyEa2I+KSK7lcH+jtj9f2urjtX99EAAMAw7pMb93yM5jy+7vloFq2m0Wbu9eOeP2mVCzwR8ZSkneUfFyJiPqX0St8hB1W0RQcAAFPiPrlxz0c9ruPrno9m0WoabeZeP+750zDqEa17y++hohvW9oHfL/X9vs4XAADI4D65cc9HfY7j656PZrnXD/XZbe71454/LaMe0fqmpPvLn5cknRr4/XlJGyUdLX+uMi9px2o/IAAAKLhPbtzz0Sz3+qE+u829fmg13W1dqB/n/GmqXOBJKS1ExB4VizMHBh7PkopFn6MppQdHnSgidooFHgAA1sR9cuOej2bRjQht5l4/tJrutq7Uj2v+tI3cZDmldLDi109IWqx5rpdV3BEEAABWwX1y456PZvW6EW3YsMGyfqjPbnOvH1pNd1uX6scxvwkjF3iqpJQ+v4pjn5X0QM75AACYNe6TG/d8NItuRGgz9/pxz0c19/F1z2/KqE2WAQBAg5wnN+75aFb/+K5bN/4pK/WJHO71456Pau7j657fpIku8ETEbRFxyyTPAQBAl7lObtzz0Sz3+qE+u829ftzzUc19fN3zm5a1wBMRX46I24b8boeK/XkuRMSbEfG7OecCAGAWOU5u3PPRLPf6oRtRt3WhfpzzUc19fN3z2yD3Dp7dKjpsXSeldFzSJkm3SvqwpI+yyAMAQLPcJ0+zMDmbZe71QzeibutK/bjmo5r7+Lrnt8VEH9FKKV0sv45Juk/SI5M8HwAAGI5W02gz9/qhG1G3dal+HPMxmvP4uue3yTQ3Wd6oIXf7AACAyeq1mnadPM3S5GwWudePez6quY+vez7qcR1f9/y2qdUmvdxnZ9jizL0RUfX2ufK9+yQtreKzAQCAMaDVNNrMvX7c81HNfXzd81Gf4/i657dRrQUeSY9KelhSWuF3e8qvOo7WPA4AAIxB/+RmEpuzMjlDDvf6cc9HNffxdc9Hs9zrZ1brs9YjWimlR1RsmPxhSX8sqfKWnQHRd/z+VX06AACwZu6Tp1mdnM0K9/pxz0c19/F1z0ez2LPPV+09eHqbJaeUFiRtV7Fok3R1AWfY15KKO3e2p5SeG+/HBwAAK3GfPNFqutu6UD/O+ajmPr7u+WgWe/Z5q/uI1jVSSmci4j5JT0namVL6s/F+LAAAsFbukydaTXdbV+rHNR/V3MfXPR/NYs8+f2vuolW2Pj83xs8CAAAyuU+eaDXdbV2qH8d8jOY8vu75aFb/+K5bN/5m29TndOSO3G5Jp+ocGBG3lN24AADABLhPntzzUc19fN3zUY/r+Lrno1nu9UN9XpW1wJNSOp5SeqXm4bslHck5HwAAWJn75Mk9H9Xcx9c9H/U5jq97PprlXj/s2Xet8d97NdwWSXNTPB8AADOhC5Mn53xUcx9f93w0y71+qM9uc68f9uy73po2WR4UEfdIWpA0P+SQWyVtlXRmHOcDAACFrkyeXPNRzX183fPRLFpNo83c64c9+1aWvcATEQ9J+sqow1S0VF/MPR8AACh0afLkmI9q7uPrno9m9VpNb9iwwbJ+qM9uc68f9/xJylrgiYiNkh5XsXgTIw6/KGl/zvkAAEDBfXLjno/RnMfXPR/NotU02sy9ftzzJy33Dp595feLkp6UtCTpYUmH+o7ZquLRrV0ppecyzwcAwMxzn9y456Me1/F1z0ez+sd3EpuzUp/I4V4/7vnTkLvAs1PS0ymlD/deiIitkh7r764VEXOSno6IhZTSi5nnBABgZrlPbtzzUZ/j+Lrno1mD4zvuBR7qEznc68c9f1pyu2jNSdo78NqSpO39L6SUliQdWOFYAABQk/vkxj0fzXKvH+qz29zrh1bT3daF+nHOn6bcBZ7NKzx2dV7SvYMHppSOqrjjBwAArJL75MY9H82iGxHazL1+aDXdbV2pH9f8actd4ImIuGXgtdOSdq1w4EYNb6MOAACGcJ/cuOejWb1uRK71Q312m3v90Gq627pUP475Tchd4FmS9EBEfDki7ihfOyZpS0Q8NnDsA+XxAACgJvfJjXs+mkU3IrSZe/2456Oa+/i65zcld5PlUyrapEvSDkk/n1I6FxFLkvZGxLzKBR9JeySdyTwfAAAzxXly456PZtGNCG3mXj/u+ajmPr7u+U3KvYPnoKQov97R9/rnytcWVCwA7SlfP5Z5PgAAZorr5MY9H81yrx/qs9vc68c9H9Xcx9c9v2lZCzwppWOSPq/izpxH+14/KOmcikUeld+XJO3POR8AALPGcXLjno9mudcP3Yi6rQv145yPau7j657fBrmPaCmlNKz1+VZJX1Xx6NZ5SbtTSq/kng8AAKyd++RpFiZns8y9fuhG1G1dqR/XfFRzH1/3/LbIXuAZJqV0UcUjWgAAoAVoNY02c68fuhF1W5fqxzEfozmPr3t+m+TuwQMAAAzQahpt5l4/7vmo5j6+7vmox3V83fPbZqILPBFxW0TcMslzAACAarSaRpu51497Pqq5j697PupzHF/3/DbKWuCJiC9HxG1DfrdD0qKkCxHxZkT8bs65AADA6vVPbtatG/9/12Fyhhzu9eOej2ru4+uej2a518+s1mfuTG+3pPmVfpFSOi5pk6RbJX1Y0kdZ5AEAYHrcJ0+zOjmbFe71456Pau7j656PZrFnn6+JPqKVUrpYfh2TdJ+kRyZ5PgAAUHCfPNFqutu6UD/O+ajmPr7u+WgWe/Z5m+Ymyxs15G4fAAAwPu6TJ1pNd1tX6sc1H9Xcx9c9H81izz5/tdqkl/vsDFucuTciqt4+V753n6SlVXw2AACwSu6TJ1pNd1uX6scxH6M5j697PprVP76TuMuV+pyOWgs8kh6V9LCktMLv9pRfdRytedxYRMS8pJ2SjqWUFlfxvp2STqWUWJACANhwnzy556Oa+/i656Me1/F1z0ezBsd33As81Of01HpEK6X0iIoNkz8s6Y8lVd6yMyD6jt+/qk+XISIOqFjceVLSzoh4uOb7dkp6epKfDQCAcXOfPLnno5r7+Lrnoz7H8XXPR7Pc64c9+65Vew+e3mbJKaUFSdtVLNokXV3AGfa1pOLOne0ppefG+/FXFhF7JM2llA6llJZSSockbYmIx0e8r7e4s5e7dwAALroweXLORzX38XXPR7Pc64f67Db3+mHPvuvVfUTrGimlMxFxn6SnJO1MKf3ZeD/WtfruxpkrXzojaX9K6cyQt+yTtKP/hZTS3og4UmadVfEI1pkyf6ukB1XsFbSYUjo4gf8ZAACMXVcmT675qOY+vu75aBatptFm7vXDnn0rW3MXrbL1+bkxfpbrRMR8RByR9ERKaVtKaYukbeWvT5eLNYPv2aXi7p2VFn8WU0p7VTy2NR8Re8rjpeLxsa2SFsb/vwQAgPHr0uTJMR/V3MfXPR/NotU02sy9ftzzJ2lNd/D02S3p9Dg+yBAHJD3U/7hU+fNCufCzJyLOlo9g9cxrRLeuMuOaDZ/Lx7eOVdwVBABAa7hPbtzzMZrz+Lrno1m0mkabudePe/6krfkOHklKKR1PKV0c14fp19sPp2IvnIfK74P76myRNKxj1tmImBt8sXxE64GU0u41fVgAAKbIfXLjno96XMfXPR/N6h/fdeuy/qo1Mp/6xGq51497/jRkXXUi4o6IuKf8uqPv9dsi4mREvFl+fTsibl9lfOWjUuXCT28PnZ19v1rS1b16Bs0NWTA6oqsLRgAAtJb75MY9H/U5jq97PprlXj/UZ7e51497/rTkPqJ1VNJmFd2yzkp6X0RsVLHwslFX26Nvl3Q8IuZTSq/UzJ6X9HBEbCn3zVnJoop9c+b7Xjs78Od+7xh8oey4dSaldHSF4wEAaA33yY17PprlXj/UZ7e51w+tprutC/XjnD9NufcNPi7poqR7U0rvK1/bp6t30BxJKa2TdKukFyU9uorsp1XcjXOy4pjeefofyXpSklZ6FGtQRMyXn5e7dwAAreY+uXHPR7PoRoQ2c68fWk13W1fqxzV/2nIXeHZK+khK6Xjfa3v6fn5Ieutxqkcl3Vs3OKV0MKW0acSdNdvL76f63rck6ZCKhZu3lN2ynhh4/xFJeyv2+QEAoHHukxv3fDSLbkRoM/f6odV0t3Wpfhzzm5C7wLNd0rHeHyJic/ljUtGRqv9xrJMqHqcai3LfnTlJRwcXaMrNkud7LdDL77f2d8iKiIfLY/s7cAEA0Crukxv3fDSLbkRoM/f6cc9HNffxdc9vSu4ePBpYxOnf7PjMwHEXI0JjtHfg++DnWoiIneUeO8f67wQqH986IGlb/3vK13dKWkopHdMI5SLRw5L0rne9SydOnFjL/46Z9uqrr/LPDZaoXUzLj3/8Y61fv17Ly8vZexj0HnPp//Ply5fHlr/S+dqcz/+HJyv3kYz+8R1HXlV+G+uzCrU7OXXrrM74Dl5zV8O5PoflU7eTNc5/v4+q3S7WZxWn2s1d4BlcselfMLlm75zy7p6xtFQv78jZKWkhpTSsJbrKRZqVFmoOSzrU/96yVfoBSbslzUXEgYrNnXv5h1Q8Dqbt27enu+++e7X/U2beiRMnxD83OKJ2MS3PPvvs2P7LUv9t8r3/crVhw4aJ/pexNufz/+HJeuGFF9b83sHxHfcjHg71WYXanZw6dVt3fNdat+71OSyfup2sXu2OY3yrarer9VnFqXZzF3gWI+L2lNLz5Z+H3sGjYuFk5F0xo5R32RxWsXfOqjtflYtDW1NKC32vzUk6Lmlz73GviFiss8gDAMAkOd6W7J6PZrnXD92Iuq0L9eOcj2ru4+ue3wa5CzzHJX01Ivaq2EB5XsX+O2dSSi/2DoqI+yV9StKWzPP1zrk/pXRwje8/LGnHwGv7VDzG9dZePin9/+zdf7Rdd33e+ed7MapNIlsSaUmctgEpCSSeVUAW6dBmMgyWySQTGqaxRGbaADMdZGiBpm1iA50fbdoAMm3TEDJBJm0DNJn6RzJNkzYlFiymSctqLAuzBtNAa5kkgyFNLV/j1HYk+37mj72PdHR17z773H3O/uxnn/drLS3bV0fPOV77c3T3fc7e32+sl1IeKaXsYRFmAMBYuJ88rcLJ2Spznx92Ixq3scyPaz6auR9f9/yh6LrI8rskfaOqLc2nd886IkmllBtKKfeq3rpc0vd1ebJSyj2S7thpuVNKOSHpzunFlms3aevt2O+WdHQnzwUAwNCw1TSGzH1+2I1o3MY0P475mM35+LrnD0mngiciHlO1M9YvSHpI1S1Yh6au3jkiaW/9ew9J+v6dPlddztzTodw5KOlovcPWZvslXbaWT71GzyKuOgIAIBVbTWPI3OfHPR/N3I+vez7acT2+7vlDs4hdtB5SfcXOFr/3pq75klTvhLW+VblTr59zqMWuV3dJeuM2f16SuA0LADBKbDWNIXOfH/d8NHM/vu75aM/x+LrnD1HXW7SWrl4U+bkNix0f3ubr0xnHJZ3ZalHmqfV19mz+PQAA3E2f3KytLf7bPidn6MJ9ftzz0cz9+LrnI5f7/KzqfA664Klvq3rZjJ2sXibpVEPGfknHtM1VRrV1VbdpbfVnH2z3agEAGBb3k6dVPTlbFe7z456PZu7H1z0fuVizz1fnW7SWpS5XPibpVL248lb2S9o/owC6S9WW6k23YN2uqija7Kb69wAAsOJ+8sRW0+M2hvlxzkcz9+Prno9ckzX7du/ebTk/qz6fgy14JJ1QddvUrFuwNu+IdUEp5ZgkRcSskubdkh6a3hK9XpvnuWyRDgBw437yxFbT4zaW+XHNRzP34+uej1ys2edvsAVPRNzY5c/XBc1xSde3eK71UsoNku4qpUx22bq1/gUAgA33kye2mh63Mc2PYz5mcz6+7vnINX18l3GVK/PZj8EWPAtypN7qfKaIOF1KOaLqiqH1bbZTBwBgsNxPntzz0cz9+Lrnox3X4+uej1ybj++iCx7msz+jLXjqW6tmbZ2+1Z+5bKctAACGzv3kyT0fzdyPr3s+2nM8vu75yOU+P6zZd6lB76IFAABmG8PJk3M+mrkfX/d85HKfH+Zz3NznhzX7LkfBAwCAsbGcPLnmo5n78XXPRy62msaQuc8Pa/ZtjYIHAABTYzp5csxHM/fj656PXJOtpl3nh/kcN/f5cc9fJgoeAAAMuZ/cuOdjNufj656PXGw1jSFznx/3/GWj4AEAwIz7yY17PtpxPb7u+cg1fXzX1hb/oxbziS7c58c9vw+9FTyllBeUUt7d1/MBADBG7ic37vloz/H4uucjl/v8MJ/j5j4/7vl96fMKnv2Sbunx+QAAGBX3kxv3fORynx/mc9zc54etpsdtDPPjnN+nK5p+s5TyQwt8rm9bYBYAACvF/eTGPR+52I0IQ+Y+P2w1PW5jmR/X/L41FjySbpMUfbwQAACwNfeTG/d85JrsRrR7927L+WE+x819fthqetzGND+O+RlmFTzrkq6RVBb0fJRFAADMwf3kxj0fudiNCEPmPj/u+Wjmfnzd87PMKnjOqip47q7/vYv9km7omAEAwEpxPrlxz0eu6eO7jPU7mE904T4/7vlo5n583fMztbmC5+6IeG3XJyqlHBYFDwAAc3E9uXHPR67Nx3fRBQ/ziS7c58c9H83cj697frZZBc9JSf9hQc/1iKSfX1AWAAArwfHkxj0fudznh92Ixm0M8+Ocj2bux9c9fwgaC56IePsCn+tBsU06AACp3E+eVuHkbJW5zw+7EY3bWObHNR/N3I+ve/5QrPX4XDdLuqvH5wMAAFPYahpD5j4/7EY0bmOaH8d8zOZ8fN3zh6TPgueApD09Ph8AAKhNtpp2PXlapZOzVeQ+P+75aOZ+fN3z0Y7r8XXPH5pZa/C0Ukp5paQjqnbK2so+SQclnV7E8wEAgPbYahpD5j4/7vlo5n583fPRnuPxdc8fos4FTynljZI+MOthkkLSma7PBwAA2mOraQyZ+/y456OZ+/F1z0cu9/lZ1fnsdItWKeUaSScm/9nwS5Iek/TuLs8HAADacz95WtWTs1XhPj/u+Wjmfnzd85GLNft8db2C5x31Px+TdKekdUnHJN0+9ZiDqm7duiki7u/4fAAAoAX3kye2mh63McyPcz6auR9f93zkmqzZt3v3bsv5WfX57FrwHJZ0T0R85+QLpZSDkt4VEV+Z+toeSfeUUo5ExBc6PicAAGjgfvLEVtPjNpb5cc1HM/fj656PXKzZ56/rLlp7JN266Wvrkg5NfyEi1iUd3+KxAABggdxPnthqetzGND+O+ZjN+fi65yPX9PFdW1v8ZtvMZz+6XsHzgi1uuzor6UZJH5/+YkTcXUphDR4AAJbE/eTJPR/N3I+vez7acT2+7vnZ/smbfzH7JXR2xfruHf/ZiFCJ0JVlr0op+kMbobW1MvsPtvD0n36c+exR12qulFKu3vS1+yTdtMUDr9H226gDAIAO3E+e3PPRzP34uuejPcfj656PXBGhjQitlaJSFlPqTHOfT7c1+7oWPOuSjpZSfqqU8pL6ayclHSilvGvTY4/WjwcAAAs0hpMn53w0cz++7vnI5T4/zOe4LbvciQjr+XRcs6/rLVqndHGb9BskfXNEPFRKWZd0ayllv+rCR9Itkk53fD4AADDF/eTePR/N3I+vez5ysdU0hqyPcmcjwnY+Xdfs63oFz22SSv3ruVNff0/9tSOqCqBb6q+f7Ph8AACg5n5y756PZu7H1z0fuSZbTbvOD/M5bn2VO2ulWM6n8/x3Kngi4qSk96q6MuftU1+/TdJDqkoe1f9cl8QiywAALID7yY17PmZzPr7u+cjFUDDjxQAAIABJREFUVtMYsj7LHdb06V/XW7QUEdttfX5Q0k+runXrrKSbI+IrXZ8PAIBV535y456PdlyPr3s+ck0f32Uszsp8ogvKndz8PnQueLYTEY+pukULAAAsiPvJjXs+2nM8vu75yLX5+C664GE+0QXlTm5+X7quwQMAAHrifnLjno9c7vPDfI6b+/yw1fS4Ue7k5vept4KnlHJDKeWRvp4PAIAxcT+5cc9HLnYjwpC5zw9bTY8b5U5uft/6voJnT8/PBwCAPfeTG/d85GI3IgyZ+/yw1fS4Ue7k5mfovAZPKeX5km6VdEjNBc6+rs8FAMCqcT+5cc9HLnYjwpC5z497PppR7uTmZ+lU8JRSXiDpP7R9uKTo8nwAAKwa55Mb93zkYjciDJn7/LjnoxnlTm5+pq63aB1XVdy0+QUAAObkenLjno9c7vPDfI6b+/y456MZ5U5ufraut2gdlLQu6XZJD8547AFJP9zx+QAAWCmOJzfu+cjlPj/sRjRuY5gf53w0o9zJzR+CrgXPfkn7I+ILsx5YSrlB0i0dnw8AAHTgfvK0Cidnq8x9ftiNaNzGMj+u+WhGuZObPxRdb9Fal3S25WPPqFqMGQAAJGCraQyZ+/ywG9G4jWl+HPMxG+VOXv6QdC14TqraPWumiHgoIt7b8fkAAMAOsNU0hsx9ftzz0cz9+Lrnox3KnZz8oela8LxH1ULLM5VSXlBK+amOzwcAAObEVtMYMvf5cc9HM/fj656P9ih3+s8fok4FT0SclvTBUspHSym7Zzx8v6RjXZ4PAADMZ/rkZm2t6+c6zfmcnGFe7vPjno9m7sfXPR+5KHc8dVpkuZTySkn/QdKnJK2XUk5LOqVqbZ7NDnR5LgAAMJ/NJzeL3n2HkzN04T4/7vlo5n583fORLbQRy7vti/lcnq67aJ2UFPW/F1Xbph/c5rFl6rEAAGCJ3E+e2Gp63MYwP875aOZ+fN3zkSsiFBFaW1uj3DG0iF20Sv1LU/++1S8AANAD95Mntpoet7HMj2s+mrkfX/d85JrcllW4csdW1yt4zkrao+pKnq1uy5q2X9JLOz4fAABo4H7yxFbT4zam+XHMx2zOx9c9H7mm19yJJdx3w3z2o2vBsy7pRES8edYDSymHJX204/MBAIBtuJ88ueejmfvxdc9HO67H1z0fuTYvqBwLbniYz/50vUXrlKT7Wj72EUkf6/h8AABgC+4nT+75aOZ+fN3z0Z7j8XXPRy52y5qd76TTFTwR8aY5HvspSa/q8nwAAOByYzh5cs5HM/fj656PXO7zw3yO27LLnYiwnk/HNfu6XsEDAAASuZ/cu+ejmfvxdc9Hro2NDev5YT7HrY9yZyPCdj5d1+zrugbPBaWUqyUdk3SjpH2Szki6V9LdEfGFRT0PAACouJ/cu+ejmfvxdc9HrvPnz+vcuXPavXu35fwwn+PWV7mzVorlfDrP/0IKnlLKGyV9YNOXD0q6SdLxUsotEfF3F/FcAADA/+TGPR+zOR9f93zkmhzfXbt2Wc4P8zlufZY7rOnTv863aJVSvk8Xy52y6dfka7eVUv5a1+cCAAD+Jzfu+WjH9fi65yPX9PFdW1v8ahjMJ7qg3MnN70Onv3VKKddIuksXC50zkk5P/Xps8lBVJc/VXZ4PAIBV535y456P9hyPr3s+crnPD/M5bpQ7ufl96VorH6v/eYukvRHxjRFxaOrXPkl7Jb1XVcnz9o7PBwDAynI/uXHPRy73+WE+x819fthqetwod3Lz+9S14HmtpBMR8Xci4rGtHhARj0XErZI+qGoBZgAAMCf3kxv3fORiNyIMmfv8sNX0uFHu5Ob3rWvB8wJJx1s+9oSk/R2fDwCAleN+cuOej1yT3Yhc54f5HDf3+WGr6XGj3MnNz9B1F609c2yB/qCkPR2fDwCG64G7pTP3SA+fkh49Iz21Lu3dX/3af6P07bdkv0IYcj+5cc9HLnYjwpC5z497PppR7uTmZ+la8DxWSnl+y5JnX8fnAoBhOnW7dPLWqtCRpK87KF13VNp7QHryEelLp6vfv++E9AP3SPu4mBHtOZ/cuOcj1/TxXcb6HcwnunCfH/d8NKPcyc3P1LXgOSPp+yT93RaPval+PACMw5Pr0l1HpDMnq//eu186cpd07cFLH3f2TPW4L52u/nnzff2/VthyPblxz0euzcd30QUP84ku3OfHPR/NKHdy87N1LXg+JumdpZSTEfHp7R5USrlB0nsk3d7x+QBgOG6/vroVS6qu2tmquHlyXXrfgYv//aXTVeHDVTxoyfHkxj0fudznh92Ixm0M8+Ocj2aUO7n5Q9C14PmApB+WdLqUckbSSUnrU7+/X9JBXVxc+UTH5wOAQfgTn/6hi+WOVF25s5WHT13+tUcpeJDH/eRpFU7OVpn7/LAb0biNZX5c89GMcic3fyg6FTwR8VAp5YOS3qiqxDm2xcMmR/+uiLi/y/MBwCA8cLf2PTp1tc7+w9sXNnu3+PpWXwN6wFbTGDL3+WE3onEb0/w45mM2yp28/CHpuk26IuJmST+vqsiZ/JqY/Ps9EfHars8FAINw8tZL//tbj2z/2H37pe85IV25p/r1PSfaXb1z5xHp4dPdXucDd1c5gNhqGsPmPj/u+Wjmfnzd89EO5U5O/tB0vUVLkhQRR0oph1VdwTN9S9ZpSSci4oOLeB4ASPfw6UtvzZKqK3iaHDpW/ZrHl05Xa/z8wD3SgRn5W3ng7mpBZ64WgthqGsPmPj/u+Wjmfnzd89Ee5U7/+UPU+QqeiYg4GRFHI+IbI2Kt/nWIcgfAqDxwx+Vfu2rf4p/n2H1VOfORG6UHT873Zx88ebHcOcaOXatu+uRmbW1h3/a3zOfkDPNynx/3fDRzP77u+chFueNp8Wd62yilXFNK+bN9PR8ALMWZLcqWq/ZU/zx1u3TieulvlOrXjx+obpGat6CZZP7APdVtXR+5sf3tWg+erB4/KXcmrw0ryf3kaVVPzlaF+/y456OZ+/F1z0c2yh1XvRU8kg5J2mabGQAwsfn2rIkP31itzbP/cLWj1uHjVTnz2burwuXHD8y/ps6+uqS5ck91u9asP0+5gynuJ09sNT1uY5gf53w0cz++7vnIFREKyh1bM9fgKaX8kKQDko5HxBe2+L22vm2+lwYAA/TU+qX/feUe6ddvk548K/3lhy4tVb79luoKns/eXRVDO1lTZ1Ly3H599evYfdK1By9/3MOnq3LnyvrKH8qdleZ+8sRW0+M2lvlxzUcz9+Prno9ck9uyCuWOrcaCp5TyUUmTn0SOlFL2R8RXph5ym6RY1osDgEF5cv3yrz21Xl25c+ujW5cqr/5gVfBMfOTG7R+7nemS58M3SK/72KUlz8P1gsxX7qke12aXLoyW+8kTW02P25jmxzEfszkfX/d85JpecyeW8BM+89mPWbdo3Vj/s0jaq+o2q2nrU7/f5hcAjM+33rR9YXNVvT36tM3brLexb39V7EhVyTO5XYtyB1PcT57c89HM/fi656Md1+Prno9cLKg8HrNu0fp5Sd9X//u6pFObfv+spGsk3V3/e5P9km6Y9wUCwOBd99rm37/20KWLM993u/TqE/M/z7UHq5JnciXPqz8o/dIbKXcgyf/kyT0fzdyPr3s+2nM8vu75yEW5MzvfSWPBExFHSim3qCpnjm+6PUuqSp+7I2LGTzdSKeWwKHgAONvuKp29M4qVzVfwSNWCyPOsxTNx7cGLt2vddYRyB5LGcfLknI9m7sfXPR+53OeH+Ry3ZZc7EWE9n45r9s1cZDkibmv47TskbbOlzGUeUXVFEACMy6yC56p9l39tu9242rpyz+ULPmMluZ/cu+ejmfvxdc9Hro2NDev5YT7HrY9yZyPCdj5d1+zrtE16RLw3IlqVNhHxqYg42uX5ACDd122xg9VOdqzaaTlz9kx1e5Ykfc+JKuf266uvY+W4n9y756OZ+/F1z0eu8+fP69y5c7bzw3yOW1/lzloplvPpPP+dCh4AWDnXbl5rfodmXfWzlbP1VutPrVdr8Rw6Vm2JPil5ttrlC6PlfnLjno/ZnI+vez5yTY7vrl27LOeH+Ry3Pssd1vTp31ILnlLK80spVy/zOQCgV9965PKvzSpWntxiDfp5C57pcufYfRe3ST9wWDpyFyXPinE/uXHPRzuux9c9H7mmj+/a2uJ/1GI+0QXlTm5+Hzr9rVNK+alSyvO3+b0bVK3P82gp5ZlSyl/r8lwAMAgHDl++aPLDmzcY3GTz7VhX7rlY0LSxXbkzcd1NVcnz6BlKnhXgfnLjno/2HI+vez5yuc8P8zlulDu5+X3pWivfrGqHrctExMck7ZW0T9J3SnozJQ+AUTh8/NL//tLp5sdvLoCuP9b+uZ5clz5y4/blzgQlz0pwP7lxz0cu9/lhPsfNfX7YanrcKHdy8/u01Fu0IuKx+tdJSa+S9KZlPh8A9OLQMT155bUX//u+E9s/9uHTl17Bc+Ue6cbj2z9+2pP1bVePnqnW2pl11Q8lz6i5n9y45yMXuxFhyNznh62mx41yJze/b30usnyNtrnaBwDc3Hf9iYvr6Dx6RvrwjZcXKk+uS3dNrdlz5Z7qKpw2Npc7Bw63+3PX3VTtrvVovdsWJc8ouJ/cuOcjF7sRYcjc54etpseNcic3P8MVbR5Ur7OzXTlz44yDuaf+s++QxE8aAEbh6Wd/dVXW3HVEOnOy+nV8r7T/cFX8PHqm+trE/nox5LZbqn/4hvnLnYlD9S1gv3xzlXNzy1IJg+R+cuOej1zsRoQhc58f93w0o9zJzc/SquCR9HZJxyTFFr93S/2rjbtbPg4Ahu+qPdLr7qluw7qvvmrm4VNVsXPlHunrDlbFznWvnW9RZUm6at/Oyp2JScnz2bt29ucxGM4nN+75yDV9fJexfgfziS7c58c9H80od3LzM7UqeCLiTaWUWyW9TNXCyt+nrcuerUyOeEh699yvEACG7tqD0rUN6/DsxOvu6Z5x6NjFoge2XE9u3PORa/PxXXTBw3yiC/f5cc9HM8qd3PxsrdfgmSyWHBFHJB1SVdxE/c+mX+uqrtw5FBH3L/blAwAwbo4nN+75yOU+P+xGNG5jmB/nfDSj3MnNH4K2t2hdIiJOl1JeJemjkg5HxMcX+7IAAMAyuJ88rcLJ2Spznx92Ixq3scyPaz6aUe7k5g/FjnfRqrc+f2iBrwVo9PiT5/V9f/9f6cZ3f0xfPMvJDQDMi62mMWTu88NuROM2pvlxzMdslDt5+UOyoyt4ptwsie1Z0It33vlpffHRJyVJjz/1dPKrAQAvk62md+/ebXnytEonZ6vIfX7c89HM/fi656Mdyp2c/KHpVPBExMcW9UKAJu//1c/p3jOPZL8MALDEVtMYMvf5cc9HM/fj656P9ih3+s8foq5X8KiU8kpJN0q6Z6u1eEopvyrpVES8s+tzYTV9/IEv6x//6y9kvwyY+9Ef/dGF/uX+xBNP6DOf+Ywk/29Oi8h/y1vesuBXhUVhq2kMmfv8uOejmfvxdc9HLsodT50KnlLKCySdVLWb1g9vk3ebpA+UUvZExF/s8nxYPV88+4Teeeens18GRsD1m4d7PnKx1TSGzH1+3PPRzP34uucjW2gjlnfbF/O5PDteZLl2a/3PIqmUUq7e/IB6a/VvlPSqUsp/0/H5sGLe9uFT2n1l5wvNAMtvHu75yOU+P2w1PW5jmB/nfDRzP77u+cgVEQqu3LHVteA5rOoKnnVJt0fEVxoe+/b6F9DKWz90Sl989En96NGXdMp55x336zcfbhrN2T7+wJf1zjvu75SBcWE3IgyZ+/yw1fS4jWV+XPPRzP34uucj1+S2rEK5Y6trwbNf0rGI2BcRb57x2PskHer4fFgRH/n1h3TvmUf0rqMv1rdce9mFYXP53Je+ojec+KR+48GdLdL88Qe+rHfe+Wl97kvdSiKMx2Q3ItdvTnzzGzf3+WGr6XEb0/w45mM25+Prno9c02vuVDfoLBbz2Y+uBY8knZ3jcXsW8HwYud948BH95D2f15//08/XK6/72s55P3Pzy/X1e6/S2z58au6S5zcefETvvPPT+vq9V+lnbn5559cCf+xGhCFznx/3fDRzP77u+WjH9fi65yMXCyqPR9eCZ13SC1o+9pCkMx2fDyP3xbNP6K/feb9etv+5esurXriQzN1XPVvve90h7b7yCr3tw6da3671Gw8+ord9+NSFcmf3Vav9lwUu/eaxtraIfnz7fL75YV7u8+Oej2bux9c9H+05Hl/3fOSi3Jmd76TrTyhnJL2j5WOPi4IHM7ztw6ckSe86+uKF5n79vudUJc2VV+gNJz45s+Sh3MFmY/jmxMnZeLnPj3s+mrkfX/d85HKfH+Zz3JZd7kSE9Xw6rtnXteD5mKQjpZT/s5Ty/K0eUEp5SSnlXkkvlXS64/NhxN55x/364qNP6ide/7KlFCptS57ffPgrF3bvet/rDlHuwP7kiU82xm0M8+Ocj2bux9c9H7nYkAFD1ke5sxFhO5+ua/Z1LXjepWoFppslPVhKeaSUcm8p5aP1P59Rtbjy9fXjT3R8PozUPz31O/r4Z39Xb3/1t+pFHRdVbjJd8rz1Q/deVvL85sPVgsy7r7yiWrtnn9cbGovnfvLEJxvjNpb5cc1HM/fj656PXGzIgCHrq9xZK8VyPp3nv1PBExGPqdr6fDIVeyUdVLV9+sH665Pfuy0ivtDl+TBOv/nwV/SeX/qsXnP9H9VrDv2xpT/f1+97jn7i9S+TpEtKHsodbOb+zYNPNsZtTPPjmI/ZnI+vez5ysSEDhqzPcoc1ffrXeZXQiLhN0nt1scjZ/E9JOh4RbdfqwQp58nzorR+6Vy/8uqv19j9zXW/P+6Jrr9ZPvP5levypp/XWD92rjz/wZb31Q/dS7uAC928e7vlo5n583fPRjuvxdc9HLjZkwJBR7uTm92Ehf+tExK2SDki6TdJJSQ+pWm/ndkkHKHewnZ++v1q74/2vP9T7c7/o2qv1Mze/XI8/9bTeeeenJYlyB5L8v3m456OZ+/F1z0d7jsfXPR+53OeH+Rw3yp3c/L5csaigiHhI1e1aWyqlXC1pH7dpYeI9/+wBffH3Q+973UtSFzLefeUVevypp9OeH8Pi/s3DPR/N3I+vez5yuc8P8zlu7vPDhgzjRrmTm9+nxV83uL2bJd3V4/NhwD7+wJf1T+/7//Rd+5+lbzvw3JTX8MWzT+itH7pXkvT2V3+rHn/qab3hxCf1xbMsGLuq3L95uOejmfvxdc9HLnYjwpC5zw8bMowb5U5uft8WdgVPCwck7enx+TBgH/q1hyRJv3LmGf3K//HRuf/8G0588pL/vuftr5zrKqAvnn1CbzjxST3+1NP6mZtfrhdde7Wu3fscve3Dp/SGE5/UL/zgd7A9+opx/+bhno9m7sfXPR+5JrsR7d6923J+mM9xc58fNmQYN8qd3PwMCyl4SimvlHRE0v5tHrJP1a5apxfxfPD3sv37dPVVz9ajj57V3r37Gh/78KNP6IuPPnnJ1174dVfr6h0WMFuVO5L0bQeeq3cdfbHeeeen9YYTn6y2U6fkWQnu3zzc89HM/fi65yMXuxFhyNznxz0fzSh3cvOzdC54SilvlPSBWQ+TFJLOdH0+jMNbXvVCSdInPvEJveIVzQssf+TXH9JP3vP5S772jj9z3YViZh7blTsTr7zua/Wuo6LkWSHu3zzc8zGb8/F1z0eu6eO7jPU7mE904T4/7vloRrmTm5+p0xo8pZRrJJ2Y/GfDL0l6TNK7uzwf0MXjT57X2z58attyZ6IqeV6sLz76ZFUGPcmicGPl/s3DPR/tuB5f93zkcp8f5nPc3OfHPR/NKHdy87N1XWR5sv35Y6q2RL9N0nr9z8mvk6qu3HllRNzf8fmAHXn8yfPVAsqPPqn3ve7QzKt/KHnGz/2bh3s+2nM8vu75yOU+P+xGNG5jmB/nfDSj3MnNH4Kut2gdlnRPRHzn5AullIOS3hURX5n62h5J95RSjrBNOvq2udxpu2vXK6/7Wr391ef1nl/6rN7yoVN6/+sPcbvWSLh/83DPRy73+WE+x819ftiNaNzGMj+u+WhGuZObPxRdC549kv6XTV9bl3RI0scnX4iI9VLKcUm3Snpzx+fEyL3/Vz+nz33p8Qv//fCjl5/IvPufPXDJIsvvOvribcuXt3zo1NzlzsRrDv0xSbpQ8nzoTS+f689jeNy/ebjnIxdbTWPI3OeH3YjGbUzz45iP2Sh38vKHpGvB84Itbrs6K+lGTRU8khQRd5dSWIMHM9175qw+96WvXPb13VdW4/r4U09v+fvbufqqZ++o3JmYlDwfe+B3d/TnMRzu3zzc85GLraYxZO7z456PZu7H1z0f7VDu5OQPTdeCp5RSrp6+HUvSfZJu0cX1eSYPvEbbb6MOXLDoq2R+4vXNu3S18ZpDf+xC0QNP7t883PORi62mMWTu8+Oej2bux9c9H+1R7vSfP0RdF1lel3S0lPJTpZSX1F87KelAKeVdmx57tH48APTO+ZuHez5yTR/ftbWu3/ab85lPzMt9ftzz0cz9+LrnIxfljqeuV/Cc0sVt0m+Q9M0R8VApZV3SraWU/aoLH1VX9Zzu+HwAsCOu3zzc85Fr8/Fd9O47zCe6cJ8f93w0cz++7vnIFtqI5d32xXwuT9eP8m6TVOpf0wucvKf+2hFVBdAt9ddPdnw+ANgRx28e7vnI5T4/bDU9bmOYH+d8NHM/vu75yBURCq7csdWp4ImIk5Leq+rKnLdPff02SQ+pKnlU/3NdEossAxgFdiPCkLnPD1tNj9tY5sc1H83cj697PnJNbssqlDu2ut6ipYi4dZvfOijpp1XdunVW0s2bFmMGAEvsRoQhc58ftpoetzHNj2M+ZnM+vu75yDW95k7E4vOZz350Lni2ExGPqbpFCyPzYz/2Ywt78zzxxBP6zGc+c8nX3N/8XfLf8pa3LPz1YLHYjQhD5j4/7vlo5n583fPRjuvxdc9Hrs0LKseCGx7msz+L304Do+f85nTPRy52I8KQuc+Pez6auR9f93y053h83fORi92yZuc76fwTSinlh0opd0xtk46Rc35zOucjl/v8MJ/j5j4/7vlo5n583fORy31+mM9xW3a5ExHW8+m4Zl+nW7RKKe+R9MP1f75U0jd3fkUYPOc3p2s+crnPD59sjNsY5sc5H83cj697PnKxIQOGrI9yZyPCdj5d1+zregXPMV3cJv2x7i8Hq8j9mxPf/MbNfX74ZGPcxjI/rvlo5n583fORa7Ihg+v8MJ/j1le5s1aK5Xw6z/8iFlm+TdJ+sQU6doBPNjBk7vPDJxvjNqb5cczHbM7H1z0fudiQAUPWZ7nDmj7963oFzylJvxERRyPiU00PLKW8tJRyb8fnw4jwyQaGzH1+3PPRzP34uuejHdfj656PXGzIgCGj3MnN70PXv3XeJOm2UsqLWzx2n6SDHZ8PI8EnGxgy9/lxz0cz9+Prno/2HI+vez5yuc8P8zlulDu5+X3pdItWRJwppdwo6QP1QbxH0mlJZ7d4+P4uz4XxmH7zLGNxVt786MJ9ftzz0cz9+LrnI5f7/DCf4+Y+P2zIMG6UO7n5feq6i9YjU/+5R9Lhbi8HY7f5zbPov+x586ML9/lxz0cz9+Prno9crNmHIXOfHzZkGDfKndz8vnVdZLmoKnZi09e2Ew2/h5Fzf3Pyyca4jWF+nPPRzP34uucj12TNvt27d1vOD/M5bu7zw4YM40a5k5ufoWvBc1ZVwdN2i/RrOj4fTLm/OflkY9zGMj+u+Wjmfnzd85GLNfswZO7z456PZpQ7uflZFrFN+l0R8dpZDyqlHJb00QU8H8y4vzn5ZGPcxjQ/jvmYzfn4uucjF2v2Ycjc58c9H80od3LzM7UqeEopV6taJHny64CkuySdqX+1EWq+fQsj5P7mdM9HM/fj656PdlyPr3s+crFmH4bMfX7c89GMcic3P1tjwVNK+VVJN0x/SdK6pFOSFBGvavtEEfExdd+WHUbc35zu+Wjmfnzd89Ge4/F1z0cu9/lhzb5xG8P8OOejGeVObv4QzLqCZ78uXnVzt6R3R8SnlvuSMAbub073fDRzP77u+cjlPj/M57i5zw9r9o3bWObHNR/NKHdy84eizS1aIem2iHjHsl8MxsH9zemej2bux9c9H7nYahpD5j4/rNk3bmOaH8d8zEa5k5c/JG1umVqn3EFb7m9O93w0cz++7vnIdf58tdW06/wwn+PmPj/u+Wjmfnzd89EO5U5O/tC0uYLn5OYvlFKer2p79G1FxP07e0lw5f7mdM9HM/fj656PXJPjy1bTGCL3+XHPRzP34+uej/Yod/rPH6I2Bc9Wu2S9XdIxVbdvbVYkPSjpmzq8Lphxf3O652M25+Prno9c08d3GYuzMp/own1+3PPRzP34uucjF+WOpza3aD2y+QsR8SZJeyV9p6THVJU6RdKbJO2NiC3LnVLKS3b+UjFU7m9O93y043p83fORy31+mM9xc58f93w0cz++7vnIRrnjqtUaPFt9MSIei4iTko7WXzoRER+MiMe2enwp5QZJ9+3sZWKo3N+c7vloz/H4uucjl/v89JGPPGOYH+d8NHM/vu75yBURCsodW20Knlnurf9JebNi3N+c7vnIxW5EGDL3+ekrHznGMj+u+Wjmfnzd85FrcltWodyx1bngmbpiZ6u1ejBS7m9O93zkYjciDJn7/PSZj/6NaX4c8zGb8/F1z0eu6TV3qtVXFov57EebRZavL6WcavG4l5ZSzjb8/sGWr2khSin7JR2WdDIiWpdPpZTDkk5FxJa3psH/zemej1zsRoQhc58f93w0cz++7vlox/X4uucj1+YFlSO22ktp55jP/rQpeI7Vv2a5reNrWZhSynFVO3ndKeloPaS3t/hzhyXdo2oBaWynCk+eAAAgAElEQVTD+c3pno9c7EaEIXOfH/d8NHM/vu75aM/x+LrnIxe7Zc3Od9LmFq3S4le0fNzSlVJukbQnIm6PiPW62DlQSjkx489Nyp1buXqnmfOb0zkfudznh/kcN/f5cc9HM/fj656PXO7zw3yO27LLnYiwnk/HNfvaXMEzKW+aLL28qW+5uisirp/x0HdIumH6CxFxaynlrqkre05FxOk696Ck10raL+lMRAzmSqShcn5zuuYjl/v88MnGuI1hfpzz0cz9+LrnIxcbMmDI+ih3NiJs59N1zb42Bc9DWswCyvskvXSeP1BK2SPpkKQbJd2ibbZsn3r8Taqu3jm9xW+fqYuePZIO11fsnKl/vVvVLmBH5nl9WAz3b0588xs39/nhk41xG8v8uOajmfvxdc9HrsmGDLt377acH+Zz3Poqd9ZKsZxP5/lvU/C8JyJ+uusT1YXKR+d8/K2STku6Q9UizYdm/LH9mlEC1bdf3b3puU6oWox5q2IIS8QnGxgy9/nhk41xG9P8OOZjNufj656PXGzIgCHrs9xhTZ/+tSl4FmWupbgj4qSkk5P/bjkcB7T91UYPllL2bF5fp75F62hEsLByz/hkA0PmPj/u+Wjmfnzd89GO6/F1z0cuNmTAkFHu5Ob3YdYiy4u4NWvirKrbvZZpXdKebX7vsnKndpekNy7vJWErfLKBIXOfH/d8NHM/vu75aM/x+LrnI5f7/DCf40a5k5vfl8YreCLiVYt6ooj4lKRvXFTeNh5UdZvWVp67+Qv1jlunI+LuLR6PJeGTDQyZ+/y456OZ+/F1z0cu9/lhPsfNfX7YkGHcKHdy8/vUZpt0J3dKFxZnblTvyvUOcfVOr9zfnGN68+Ny7vPjno9m7sfXPR+5WLMPQ+Y+P2zIMG6UO7n5fRtVwVPfgnW7quLmgnp3rTs2PfwuSbduc9sWlsD9zcknG+M2hvlxzkcz9+Prno9ckzX7XOeH+Rw39/lhQ4Zxo9zJzc8wqoJHkiLiZkn761JnUu7sm94hq5RyrH7s7TmvcvW4vzn5ZGPcxjI/rvlo5n583fORizX7MGTu8+Oej2aUO7n5WfrcRas3EXGklHK4XmPn5PQaO/XtW8clXT/9Z+qvH5a0Xu/gNVNdFB2TpOc973n6xCc+saD/g2Gbt4jY2NjQuXPntGvXLp0/f/6SK1Uml1x30ZS/CH3mr8oMZdhuznZ6fNvO7pjmsymf2V2exx9/fGHHd/PcDmV+svKZ2+Va5Pf3ReQ15Q9xPpswu8vTds7aHN8u57nO87ld/pDn9oknn8x+CZ1tbExvWh2KCJVSFFGVMfMJbWw0/37b/J28B4Y2/0Oe3c1GWfBIl2+zPuWDkm6PiAs7hNVbpR+XdLOkPaWU4xFxa4vnuF3VLWE6dOhQvOIVr1jESx+8z3zmM60fO2lGt9sKveslm7Pyu+o7f1VmKMNWc9vl+LaZ3bHNZ5Ohzu6bTx7LfgmdnX3Rowv7ZCxiQ6Ws1f8eF/57J/nf/rvf0fj7Q5rP7Qx1bsdinvOFzTYf30Xf4uEwn02Y3eVpM7dtj+9O59Z9PrfLH/LcfvmOX8x+CZ2d+4M/kDS5skZaW9vZ93dJ2tiQ1ta2/rPz5s/7Hhji/A95djcbbcGzlfp2rYMRcWTqa3skfUzSCybr8ZRSzrQtebA998vq3PPRzP34uuejvWVdVt2l3JmF+UQX7vPDmn3jNob5cc5HM27Lys0fgtGtwTPDByUd2fS1d6i6jevCYsv1vz/SZjcubM39zemej2bux9c9H7kodzBk7vPDmn3jNpb5cc1HM8qd3PyhWJmCp5RyQtKd04st126SdO8Wf+RuSUeX/sJGyP3N6Z6PZu7H1z0fuar75Cl3MEzu89NnPvo3pvlxzMdslDt5+UOyErdo1WvsHI2IvVv89n5JZzZ/MSLOlFIOLP3FjYz7m9M9H83cj697PnJFhEKhNcodDJD7/Ljno5n78XXPRzuUOzn5Q7MqV/DcJemNm784dQvW+ubfw/zc35zu+Wjmfnzd85Hrwm1Z4uQMw+M+P+75aOZ+fN3z0R7lTv/5QzT6gqeUclzSmemt0iem1t1hrZ2O3N+c7vmYzfn4uucj16Vr7iw+n/lEF+7z456PZu7H1z0fuSh3PLndojVXEVNK2S/pmKQXNDxsXdVtWlv92QfnenUryv3N6Z6PdlyPr3s+cm1eUDkiFprPfKIL9/lxz0cz9+Prno9s9VbolDt2nK7g2S9dcltVG3dJunV6h6wt3C7pZVt8/SZJd87xXCvJ/c3pno/2HI+vez5ysVvW7HzkGcP8OOejmfvxdc9Hrur8gSt3XA224CmlHCyl3Ff/elQXr7J5aOrrNzX8+WOSFBG3z3iqd0s6PF0c1f/+3BnF0Mpzf3O65yPXxsaG9fwwn+O27HInIqznc5KPHO5/v7nno5n78XXPR67JbVmFcsfWYG/Rqrczv34nf7YuaI63+fMRsV5KuUHSXaWUm+sv31r/wjbc35zu+ch1/vx5nTt3Trt377acH+Zz3PoodyI2bOdzOh/9c//7zT0fszkfX/d85Jpec2fBd3RLYj77MtiCZwGORMRl259vJSJOl1KOSDosaT0ibp71Z1aZ+5vTPR+5Jsd3165dlvPDfI5bX+VOKWuW88n853I/vu75aMf1+LrnI9fmBZVZs8/XKAue+taqkzv4M5fttIXLOb853fORa/r4LmP9DuYTXfRZ7nDZNublfnzd89Ge4/F1z0cudsuane9ksGvwYLic35zO+cjlPj/M57hR7uTmo5n78XXPRy73+WE+x23Z5Q5r9vWPggdzc35zuuYjl/v88MnGuFHu5Oajmfvxdc9HLjZkwJD1Ue5sRNjOp+uafRQ8SOf+zYlvfuPmPj98sjFulDu5+Wjmfnzd85FrsiGD6/wwn+PWV7mzVorlfDrPPwUPUvHJBobMfX74ZGPcKHdy8zGb8/F1z0cuNmTAkPVZ7nD+0D8KHqThkw0Mmfv8uOejGeVObj7acT2+7vnINX1819YW/6MW84kuKHdy8/tAwYMUfLKBIXOfH/d8NKPcyc1He47H1z0fudznh/kcN8qd3Py+UPCgd3yygSFznx/3fDSj3MnNRy73+WE+x819ftiQYdwod3Lz+0TBg165vznH9ObH5dznxz0fzSh3cvORizX7MGTu88OGDONGuZOb3zcKHvTG/c3JJxvjNob5cc5HM8qd3HzkYs0+DJn7/LAhw7hR7uTmZ6DgQS/c35x8sjFuY5kf13w0o9zJzUcu1uzDkLnPj3s+mlHu5OZnoeDB0rm/OflkY9zGND+O+ZiNcicvH7lYsw9D5j4/7vloRrmTm5+JggdL5f7mdM9HM/fj656Pdih3cvKRy31+mM9xc58f93w0o9zJzc9GwYOlcX9zuuejmfvxdc9He5Q7/ecjl/v8sGbfuI1hfpzz0YxyJzd/CCh4sBTub073fDRzP77u+chFuYMhc58f1uwbt7HMj2s+mlHu5OYPxRXZLwDj4/7mdM9HM/fj656PXBGSRLmDYXKfH9bsG7cxzY9jPmaj3MnLHxKu4MFCub853fPRzP34uucjV0QoFJQ7GCT3+XHPRzP34+uej3Yod3Lyh4YreLAw7m9O93w0cz++7vnIdeG2LHFy5uIPfvLfZr+Ezm5a/6Z2D5z64UH/eev5jNhQObfDzyU35d+959/vLGcbzP+4uR9f93y0R7nTf/4QUfBgIdzfnO75mM35+LrnI9f0mjtSLDyf+UQn0+XLEn54WHY+8z9u7sfXPR+5KHc8cYsWOnN/c7rnox3X4+uej1wsqIxBo9xJzUcz9+Prno9slDuuKHjQifub0z0f7TkeX/d85KLcmZ2PRJQ7qflo5n583fORqzp/oNxxRcGDHXN/c7rnI9fGxob1/DCf47bscicirOeTraaTUe6k5qOZ+/F1z0euyW1ZhXLHFgUPdsT9zemej1znz5/XuXPnbOeH+Ry3PsqdiA3b+WSr6WSUO6n5mM35+LrnI9f0mjsSf3+6ouDB3NzfnO75yDU5vrt27bKcH+Zz3Poqd0pZs5xP5j8Z5U5qPtpxPb7u+cjFgsrjQcGDuTm/Od3zkWv6+K6tLf6vT+YTXfRZ7nDyh7lR7qTmoz3H4+uej1yUO7PznVDwYG7Ob07nfORynx/mc9wod3LzMQPlTmo+crnPD/M5bssud1izr38UPJib85vTNR+53OeHTzbGjXInNx8zUO6k5iMXGzJgyPoodzYibOfTdc0+Ch6kc//mxDe/cXOfHz7ZGDfKndx8zEC5k5qPXGzIgCHrq9xZK8VyPp3nn4IHqfhkA0PmPj98sjFulDu5+ZiNcicvH7nYkAFD1me5w/lD/yh4kIZPNjBk7vPjno9mlDu5+WiHcicnH7nYkAFDRrmTm98HCh6k4JMNDJn7/LjnoxnlTm4+5kC503s+crnPD/M5bpQ7ufl9oeBB7/hkA0PmPj/u+WhGuZObj2SUOxgw9/lhQ4Zxo9zJze8TBQ965f7mHNObH5dznx/3fDSj3MnNR66I5a7pw3yiC/f5YUOGcaPcyc3vGwUPeuP+5uSTjXEbw/w456MZ5U5uPpJFSKLcwTC5zw8bMowb5U5ufgYKHvTC/c3JJxvjNpb5cc1HM8qd3Hwkq394kCh3MDzu8+Oej2aUO7n5WSh4sHTub04+2Ri3Mc2PYz5mo9zJy0eyS354WHw884ku3OfHPR/NKHdy8zNR8GCp3N+c7vlo5n583fPRDuVOTj6SsaAyBsx9ftzz0YxyJzc/GwUPlsb9zemej2bux9c9H+1R7vSfj2SUOzPzkWcM8+Ocj2aUO7n5Q0DBg6Vwf3O656OZ+/F1z0cuyh0M2pLLHUVYzydr9uVy//vNPR/NKHdy84fiiuwXgPFxf3O656OZ+/F1z0euCEmi3MFA9VDubETYzidr9uVy//vNPR+zUe7k5Q8JV/BgodzfnO75aOZ+fN3zkSsiFArKHQxTT+XOWimW88n853I/vu75aIdyJyd/aCh4sDDub073fDRzP77u+ch14bYscXKGAeqx3GFNH8zL/fi656M9yp3+84eIW7SwEO5vTvd8zOZ8fN3zkWt6zR0pFp7PfKITyp3UfDRzP77u+chFueOJK3jQmfub0z0f7bgeX/d85GJBZQwa5U5qPpq5H1/3fGSj3HFFwYNO3N+c7vloz/H4uucjF+XO7HwkotxJzUcz9+Prno9c1fkD5Y4rCh7smPub0z0fuTY2Nqznh/kct2WXO8FW0+iCcic1H83cj697PnJNbssqlDu2KHiwI+5vTvd85Dp//rzOnTtnOz/M57j1Ue5EbNjOJ1tNJ6PcSc3HbM7H1z0fuabX3JH4+9MVBQ/m5v7mdM9Hrsnx3bVrl+X8MJ/j1le5U8qa5Xwy/8kod1Lz0Y7r8XXPRy4WVB4PCh7MzfnN6Z6PXNPHd21t8X99Mp/oos9yh5M/zI1yJzUf7TkeX/d85KLcmZ3vhIIHc3N+czrnI5f7/DCf40a5k5uPGSh3UvORy31+mM9xW3a5w5p9/aPgwdyc35yu+cjlPj98sjFulDu5+ZiBcic1H7nYkAFD1ke5sxFhO5+ua/ZR8CCd+zcnvvmNm/v88MnGuFHu5OZjBsqd1HzkYkMGDFlf5c5aKZbz6Tz/FDxIxScbGDL3+eGTjXGj3MnNx2yUO3n5yMWGDBiyPssdzh/6R8GDNHyygSFznx/3fDSj3MnNRzuUOzn5yMWGDBgyyp3c/D5Q8CAFn2xgyNznxz0fzSh3cvMxB8qd3vORy31+mM9xo9zJze8LBQ96xycbGDL3+XHPRzPKndx8JKPcwYC5zw8bMowb5U5ufp8oeNAr9zfnmN78uJz7/LjnoxnlTm4+ckUsd00f5hNduM8PGzKMG+VObn7fKHjQG/c3J59sjNsY5sc5H80od3LzkSxCEuUOhsl9ftiQYdwod3LzM1DwoBfub04+2Ri3scyPaz6aUe7k5iNZ/cODRLmD4XGfH/d8NKPcyc3PQsGDpXN/c/LJxriNaX4c8zEb5U5ePpJd8sPD4uOZT3ThPj/u+WhGuZObn4mCB0vl/uZ0z0cz9+Prno92KHdy8pGMBZUxYO7z456PZpQ7ufnZKHiwNO5vTvd8NHM/vu75aI9yp/98JKPcmZmPPGOYH+d8NKPcyc0fAgoeLIX7m9M9H83cj697PnJR7mDQllzuKMJ6PlmzL5f732/u+WhGuZObPxRXZL8AjI/7m9M9H83cj697PnJFSBLlDgaqh3JnI8J2PlmzL5f732/u+ZiNcicvf0i4ggcL5f7mdM9HM/fj656PXBGhUFDuYJh6KnfWSrGcT+Y/l/vxdc9HO5Q7OflDQ8GDhXF/c7rno5n78XXPR64Lt2WJkzMMUI/lDmv6YF7ux9c9H+1R7vSfP0TcooWFcH9zuudjNufj656PXNNr7kix8HzmE51Q7qTmo5n78XXPRy7KHU9cwYPO3N+c7vlox/X4uucjFwsqY9Aod1Lz0cz9+LrnIxvljisKHnTi/uZ0z0d7jsfXPR+5KHdm5yMR5U5qPpq5H1/3fOSqzh8od1xR8GDH3N+c7vnItbGxYT0/zOe4LbvcCbaaRheUO6n5aOZ+fN3zkWtyW1ah3LFFwYMdcX9zuucj1/nz53Xu3Dnb+WE+x62Pcidiw3Y+2Wo6GeVOaj5mcz6+7vnINb3mjsTfn64oeDA39zenez5yTY7vrl27LOeH+Ry3vsqdUtYs55P5T0a5k5qPdlyPr3s+crGg8nhQ8GBuzm9O93zkmj6+a2uL/+uT+UQXfZY7nPxhbpQ7qfloz/H4uucjF+XO7HwnFDyYm/Ob0zkfudznh/kcN8qd3HzMQLmTmo9c7vPDfI7bsssd1uzrHwUP5ub85nTNRy73+eGTjXGj3MnNxwyUO6n5yMWGDBiyPsqdjQjb+XRds4+CB+ncvznxzW/c3OeHTzbGjXInNx8zUO6k5iMXGzJgyPoqd9ZKsZxP5/mn4EEqPtnAkLnPD59sjBvlTm4+ZqPcyctHLjZkwJD1We5w/tA/Ch6k4ZMNDJn7/LjnoxnlTm4+2qHcyclHLjZkwJBR7uTm94GCByn4ZAND5j4/7vloRrmTm485UO70no9c7vPDfI4b5U5ufl8oeNA7PtnAkLnPj3s+mlHu5OYjGeUOBsx9ftiQYdwod3Lz+0TBg165vznH9ObH5dznxz0fzSh3cvORK2K5a/own+jCfX7YkGHcKHdy8/tGwYPeuL85+WRj3MYwP875aEa5k5uPZBGSKHcwTO7zw4YM40a5k5ufgYIHvXB/c/LJxriNZX5c89GMcic3H8nqHx4kyh0Mj/v8uOejGeVObn4WCh4snfubk082xm1M8+OYj9kod/LykeySHx4WH898ogv3+XHPRzPKndz8TBQ8WCr3N6d7Ppq5H1/3fLRDuZOTj2QsqIwBc58f93w0o9zJzc9GwYOlcX9zuuejmfvxdc9He5Q7/ecjGeXOzHzkGcP8OOejGeVObv4QXJH9AjBO7m9O93w0cz++7vnIRbmDQVtyuaMI6/kc+pp9v/yD35X9Ejrbs76+7e9FhDY2NvQ1a+3+/rx6Y0Nra+0/T5+Vv37oz7XO2gp/P48b5U5u/lBQ8GDh3N+c7vlo5n583fORK0KSKHcwUD2UOxsRtvPJmn25JuXLWstyZ2j5/P08fpQ7eflDwi1aWCj3N6d7Ppq5H1/3fOSKCIWCcgfD1FO5s1aK5Xwy/7kod3Lz0Q7lTk7+0FDwYGHc35zu+Wjmfnzd85Hrwm1Z4uQMA9RjucOaPpgX5U5uPtqj3Ok/f4i4RQsL4f7mdM/HbM7H1z0fuabX3JFi4fnMJzqh3EnNRzPKndx85KLc8cQVPOjM/c3pno92XI+vez5ysaAyBo1yJzUfzSh3cvORjXLHFQUPOnF/c7rnoz3H4+uej1yUO7PzkYhyJzUfzSh3cvORqzp/oNxxRcGDHXN/c7rnI9fGxob1/DCf47bscifYahpdUO6k5qMZ5U5uPnJNbssqlDu2KHiwI+5vTvd85Dp//rzOnTtnOz/M57j1Ue5EbNjOJ1tNJ6PcSc3HbJQ7efnINb3mjsR8uqLgwdzc35zu+cg1Ob67du2ynB/mc9z6KndKWbOcT+Y/GeVOaj7aodzJyUcuFlQeDwoezM35zemej1zTx3dtbfF/fTKf6KLPcoeTP8yNcic1H+1R7vSfj1yUO7PznVDwYG7Ob07nfORynx/mc9wod3LzMQPlTmo+clHuYMiWXe6wZl//KHgwN+c3p2s+crnPD59sjBvlTm4+ZqDcSc1HPsodDFUf5c5GhO18uq7ZR8GDdO7fnPjmN27u88MnG+NGuZObjxkod1LzkevCVtOUOxigvsqdtVIs59N5/il4kIqtpjFk7vPDJxvjRrmTm4/ZKHfy8pFrclsWW01jiPosd5j//lHwIA1bTWPI3OfHPR/NKHdy89EO5U5OPnJNr7mzDMwnuqDcyc3vwxXZL2DVHH3fr2W/hM4eXf+G7iER2tgIlSKVxxfzDfB79/3WhX/nzY8u3OfHPR/NKHdy8zEHyp3e85Fr84LKEbHQfOYTXVDu5Ob3hSt40L+63FlbW8KJn3jzoxv3+XHPRzPKndx8JKPcwYAte7esMexGhDyUO7n5faLgQb+myx3e/BgY9/lxz0czyp3cfOSKWO6aPswnuuij3NnY2LCdTzZkyEW5k5vfNwoe9IdyZ2Y+8oxhfpzz0YxyJzcfySIkUe5gmPoqd9bW1iznkw0ZclHu5OZnoOBBP5Zc7mgEl63yyUYe928e7vloRrmTm49k9Q8PEuUOhqfPcoe/nzEvyp3c/CwUPFi+HsqdjY2wffPzyUYu928e7vmYjXInLx/JLvnhYfHxzCe6oNzJzUczyp3c/EwUPFiunsqdtbVi+eYf818uDtyPr3s+2qHcyclHMhZUxoBR7uTmoxnlTm5+NgoeLE+P5Q4nf5iX+/F1z0d7lDv95yMZ5c7MfOSh3MnNRzPKndz8IaDgwXJQ7qTmo5n78XXPRy7KHQzakssd1uxDF5Q7ufloRrmTmz8UV2S/AIwQ5U5qPpq5H1/3fOSKkCTKHQxUD+XORrBmH3aGcic3H7NR7uTlDwlX8GCxKHdS89HM/fi65yNXRCgUlDsYpp7KnbXCmn2YH+VObj7aodzJyR8aCh4sDuVOaj6auR9f93zkunBbljg5wwD1WO5w/oB5Ue7k5qM9yp3+84eIW7SwGJQ7qfmYzfn4uucj1/SaO1IsPJ/5RCeUO6n5aEa5k5uPXJQ7nriCB91R7qTmox3X4+uej1wsqIxBo9xJzUczyp3cfGSj3HFFwYNuKHdS89Ge4/F1z0cuyp3Z+UhEuZOaj2aUO7n5yFWdP1DuuKLgwc5R7qTmI9fGxob1/DCf47bscifYahpdUO6k5qMZ5U5uPnJNbssqlDu2WIMHO0O5k5qf4idemP0KOvlz6+sLy7rk5Oz3u8/nz+793y75b+YTXfRR7kRs2M4nW00no9xJzcdslDt5+cg1veZOLH7JPuazJ1zBg/lR7qTmI1dEKDb4ZAPD1Fe5U8qa5Xwy/8kod1Lz0Q7lTk4+crGg8nhQ8GBulDt5+ch1odxZK1rC9DOf6KTPcoeTP8yNcic1H+1R7vSfj1yUO7PznVDwYG6UOzn5yHVJucN8YmAod3LzMQPlTmo+clHuYMiWXe6wZl//KHgwP8qd3vORi3Jndj7yUO7k5mMGyp3UfOSj3MFQ9VHubETYzqfrmn0UPMhHuYMBW3a5wycb6IJyJzcfM1DupOYj14Wtpil3MEB9lTtrpVjOp/P8U/AgHeUOhqqPcic2+GQDO0O5k5uP2Sh38vKRa3JbFhsyYIj6LHeY//5R8CDPhU82ODnD8PRV7pQ1PtnA/Ch3cvPRDuVOTj5yTa+5swzMJ7qg3MnN7wMFD3LUt2UVTs4wQH2WO3zzw7wod3LzMQfKnd7zkYsFlTFklDu5+X2h4EH/ptfcWQLe/OiCcic3H80od3LzkYxyBwO27HJnDGv2IQ/lTm5+nyh40C8WVMaAUe7k5qMZ5U5uPnJFLHdNH+YTXfRR7mxsbNjOJxsy5KLcyc3vGwUP+kO5MzMfeSh3cvPRjHInNx/JIiRR7mCY+ip31tbWLOeTDRlyUe7k5meg4EE/llzuaASXrfLJRh7Kndx8NKPcyc1HsvqHB4lyB8PTZ7nD38+YF+VObn4WCh4sXw/lzgZbTWOHKHdy8zEb5U5ePpJd8sPD4uOZT3RBuZObj2aUO7n5mSh4sFw9lTtrbDWNHaDcyc1HO5Q7OflIxoLKGDDKndx8NKPcyc3PRsGD5emx3OHkD/Oi3MnNR3uUO/3nIxnlzsx85KHcyc1HM8qd3PwhoODBclDupOajGeVObj5yUe5g0JZc7rBmH7qg3MnNRzPKndz8obgi+wVghCh3UvPRjHInNx+5IiSJcgcD1UO5sxGs2YedodzJzcdslDt5+UPCFTxYLMqd1Hw0o9zJzUeuiFAoKHcwTD2VO2uFNfswP8qd3Hy0Q7mTkz80FDxYHMqd1Hw0o9zJzUeuC7dlifnEAPVY7nD+gHlR7uTmoz3Knf7zh4hbtLAYlDup+ZiNcicvH7mm19yRYuH5zCc6odxJzUczyp3cfOSi3PHEFTzojnInNR/tUO7k5CMXCypj0Ch3UvPRjHInNx/ZKHdcUfCgG8qd1Hy0R7nTfz5yUe7Mzkciyp3UfDSj3MnNR67q/IFyxxW3aGHnKHdS85Gr2oyIcgfDtOxyJ9hqGl1Q7qTmoxnlTm4+ck1uyyqUO7a4ggc7Q7mTmo9ck082KHcwRH2UOxEbtvPJVtPJKHdS8zEb5U5ePnJNr7kjMZ+uKHgwP8qd1HzkunBbFp9sYID6KndKWbOcT+Y/GeVOaj7aodzJyUcuFlQeDwoezI1yJy8fuS5Zc38exYUAACAASURBVGcJ+cwnuuiz3OHkD3Oj3EnNR3uUO/3nIxflzux8JxQ8mBvlTk4+crGgMoaMcic3HzNQ7qTmIxflDoZs2eUOa/b1b+UXWS6l3CTpRknr9Zf2SDoeEWfyXtXAUe70no9clDuz85GHcic3HzNQ7qTmIx/lDoaqj3JnI8J2Pl3X7FvpgqeUckLSvog4MvW1PZLuK6XcHBEn817dCqHcwYAtu9zhkw10QbmTm48ZKHdS85HrwlbTlDsYoL7KnbVSLOfTef5XtuCpr9w5GhF7p78eEeullJsl3VVKeUFErG+dgEWh3MFQ9VHuxEboObs959P1k42xoNzJzcdslDt5+cg1uS2LDRkwRH2WO8x//1Z5DZ7jkm7f6jemrtx5R38vZwVd+GSDkzMMT1/lTlnjkw3Mj3InNx/tUO7k5CPX9Jo7y8B8ogvKndz8PqxkwVNKOShpv6R7Gx52StKxfl7RCqpvyyqcnGGA+ix3+OaHeVHu5OZjDpQ7vecjFwsqY8god3Lz+7KSBY+kw/U/mxZSPiNpTyllfw+vZ7VMr7mzBLz50QXlTm4+mlHu5OYjGeUOBmzZ5c4Y1uxDHsqd3Pw+rWrB87L6n00Fz4P1Pw8u+bWsFhZUxoBR7uTmoxnlTm4+ckUsd00f5hNd9FHubGxs2M4nGzLkotzJze/bqhY8e6RqQeWGx0x+b9/yX86KoNyZmY88lDu5+WhGuZObj2QRkih3MEx9lTtra2uW88mGDLkod3LzM6xqwTNPabNnaa9ilSy53NEILlvlk408lDu5+WhGuZObj2T1Dw8S5Q6Gp89yh7+fMS/Kndz8LCUisl9D70opD0raHxHbTmIp5ZikE5Jui4hbGx4zWYj5hZI+t+jXugK+RtJ/yn4RwA4wu3DE3MIVswtHzC1cMbvD9w0R8Yc3f/GKjFcyFhFxu7bZah3tlFJORcSh7NcBzIvZhSPmFq6YXThibuGK2fW1qrdozeOR7BcAAAAAAADQZFULnnVJKqW0WV+naSFmAAAAAACAdKta8Ey2R29abHlS/pxd8mtZddziBlfMLhwxt3DF7MIRcwtXzK6pVV1k+bikWyRdHxGnt3nMCVULKB+IiDNbPQYAAAAAAGAIVvUKnjvqf+5veMx+SeuUOwAAAAAAYOhWsuCpr9pZl3Rjw8MOi0vTAAAAAACAgZUseGpvlHR0q4WWSyk3qSqA3t37qwKABqWUZ2e/BgAAAADDs7IFT0TcLelOSR+c/npd+ByXdCQi2EELwGCUUr5G0hdLKc9PfikAAAAABuaK7BeQKSJuLqXcVC+oPClz9ki6kbV3AAxJKWWfpM+rKqbZ3Q+DV0p5VkQ8k/06gDZKKV8l6fkR8UD2awEAYKdWchctAHBSX1n4O5L+YUT85VLKWkRsZL8uYDv1rYS/IOk9EfGvs18PMEsp5UckHZV0U0R8Jvv1AACwEyt7ixYAOJgqd75K0p8qpVwVERusxYOB+6OSbpD0kVLKn8p+MUALV0v6Zkl3lFL+RPaLAQBgJyh4AGCgSil7dbHc+c+Srpf0a6WUr46I86WUlb7NFoP2RyT9nqRvkPSPSyl/Mvn1AJcppTy7lPItpZS/LemP11/+Fkk/V0q5LvGlAQCwI9yihUEqpVwt6XkR8e+zXwuQoV5z57ckPSHpD0t6WtIzkv6QpE9J+q8i4olSyhUR8XTeKwUqpZSvlvRWSX9B0rMkfb2kUv/7Q5J+ICL+Td4rBC4qpVwj6R9I+pOS/pOkfyfpBarWp7y+/u83RMS9aS8SAIA5cQUPBqf+wfZhSa8spTwr+/UAfauv3PltSX9P0gFJ/1LVDx3PkvQHkl6q6kqe50TE01zJg2z139v3SvovJf1fkm6U9GJJh1XN8fMk/UIp5YVpLxKo1fP6oKoF6/9SRLw0Iv7HiHh5RLxM0t9QVar/mcSXCQDA3LiCB4NSn3Q9JOmEpHexVT1WUSnl2yW9UtJPRcTv1V/7FUnfKa7kwcBMXW32k5J+JiJ+c4vHvFrSH4mIf9D36wOm1VfufF7ST0h6/+Q8o5SyS1JExPn6vw9ExIP1v5fghBkAYICCB4MxtZjsz0h6Z0Q8nvuKgHzTW02XUv6FpP9WlDwYiPq2rFOSfknSrZPd3SY/EG+1VfpkFzh+aEbf6hLnl1TdkvXWiDi7xWMum1kAAFxwWT8GYdNOQffvpNypb1PZHRGPLvr1ActW/6D8oog4Nf31iHhmUtxExHdPXckjXXq7FiUPMrxF0pOSfnZS7kjVZRD1v25MP7iU8jxJ07cV/m4vrxKofKuqNXd+cJtyZ2263CmlfL+kvZJeKOkrkv4fSac5zwAADBVr8CDdpnJHko6XUr53B+uKfIOkHyulvHyhLxBYslLKcyV9WdUCn5eZXmcnIr5LF9fkuULSebEmD/L8BUm/FxH3b/6NumyMUsoVpZRvKqX8dUn3SToj6QFJny2l/B22pEaP/qyq7dA/Jl34YEj1vz+rvrJsVynlfy6lfFjSz0l6v6S3SfpfVf3d+49KKQf7f+kAAMxGwYNU9Q+2vyPpOZLOqbrtZJ+qnS1eXUqZZ0bPqzpx+99LKX981oOBIagXVH5Q1bpTv7rd4+riZrLo+Gsk/aaqv8PXxMLLSFBK2S3payX956mvXZi7eg7/kKTbJN0p6W9JulbSV9e/9kr6q5J+tpRyY48vHatrcmXZ90qXlefP1H8f3yHpPZL+fP3Yp1RdiXZe1UL3/52kf1JKOdTnCwcAoA0KHqSpy52HVC10+FdV7Zy1poslzwclfW/bkiciflvScUmvUFXyXL2Elw0sTH312m9L+n8lvTciHmt6fP0DyG5Jb5D0IlUlz7O0/e5az17m68fqKqUUVX9fF0mH61tZNLk9sJRyTSnlJZL+laQfVLWjllT9kFwk7ZqKu07V1ZdcFYGFq2d14rP1P7+/lPIa6ULJs7+U8j9I+nVV5c9eXSyDrtLFc5N1VX/f7pf0N0spX9PD/wIAAK1R8CDT35f0YUk/EhE/rqqc+S3tsOSpF+z8t5LeLen1kt405xVAQG823Zp4naTvmDWvpZQrJf1Pkj4g6U0RcZ2kj2rrLdS/KiLOl8oV9SfTwI5tLgzrQvKfq7oC86+UUt5RSjlYSvmzqnbU+kVJL9PFqzOlar2eByX9TUn/sP7zkvQtqmYbWIhSyr56bbLphby/pOqKnJdL+qlSyv9dSvlHkj6h6u/Vb1E1r6GqiPwdVVf0fLuk/0LSjZL+sarzlO+SdEM//zcA4KGU8lWllOuyX8cqYxctpKlvN/nq6asWSilvknSLqvV0NlT90HpW0hsl/eL0Ip7bZK5J+tOqFkJ8RtJrIuKfN/0ZoG9T20p/laofNq6U9LikN0m6c6sdXOpy55iqYvQHI+J9U7+33e5afyoi/qCU8npJPybpG7daWBSYpb5S4UFJL4uIz099/S2S3jf10EdVXf0w8YyqH5avkHS/qjVNfj4iHprK+DuqruKUpOsj4lNL+Z/Ayqj/jv1tST8eEX990+/9pKQ362KJM/G0qjmdfP3fqLqt8HRE/N6mjH+o6krKfyHpv5f0zKzzEwBYBaWUH5F0VNJNEfGZ7Neziri6ASnqq22emZQ7k7VFIuIDqtZr2NGVPBGxERG/JulzqsqhXyqlvGJp/yPAnOpbE39L0m9I+ryqMuYZSbtVrcNzZGqtncmf2bLcqdc3UUR8ty4uvLz5Sp4fqP/c6yh3sBP11V+fl3S3qsXAL9z2EhHvV3W1zsRuVfP8VP3fk3n856pKyPdFxEOllLWpv/d/SFXxI0l/dLn/Nxi7+urI35L0s6oWSJ58fTJvf6n+velyZ1JCStUts38/Ir49Ij46Xe5MXcX2I6rKzKsi4jzlDgBccLWkb5Z0B5so5KDgQYpNl0xP1hZZq/99xyVPfSvKFar+cplsFc2JFwah3gr9t1T9QPwqVbcTntHFOf9qSbdLOjr5YaQud27W5eXOWn11zmSB0O/WpbdrPSHpkKQPSfr+iPjlvv4/MR5T60TdL+mHVV1ppnp3rMkPzG/VxZJnMn9X1v99l6q5fXVE/EdVa/BMyvhnphZl/lz9zz9eP+/0D99AK5tuff1IRHxp6txi+jzjByT9PUn/WtVMf1nVnP8tSX8xIv5qnXfJYvURcb7+1/+oas6vKaVcybyiD6yrh6EqpTy7lPItpZS/rfr7uKpbXn+O27X6xy4rGIx6e9K1+sT/A/X50uR2remSR6WULW/XqhdLfLmkr6u/9F/XV/QA6SLi90spf0nSr9S3YX2o/sHgnZIO6GLJc0LSM6WUe1Tt5PJjkv7KpnJno858ulTbUT8dEd9VSvllSd+ti3+/vyIi/lWf/58Yh/o2ly+o+mH5gKpeJ0opz66vWnimvhozIuKt9ew9T9L1qorLM5L+5eSH4lJtQ33J7YeTRZkl/b6qtU/urb/O/eOYy9S8rqsqEk+UUg5OivD678jp84wfKqU8R9W5xWOqbhn/0lTe2tR8bvYiVVf9/FxEPLXNY4CFqW+T/Wwp5dsi4gvZrweYKKVco2r34z8p6T9J+neS/q2q89DrJd1ZSnlDRNyb9ypXC2vwYHCmf3idd02e+hvg31Z1O8t3RMSv9/nage1s9cPt1O+9Xvr/2TvvcDuq6g2/66YSOkFAadIEpUroUkMvMTSV3gREkV4VgtJ7QJoUQWwURUWqdFSCqHQUhB8lASmCdEhPvt8fa0/uvpNzzj23nFvX+zw895xpZ4bsmdn722t9ixPwQXTRzj/Fy6bvwOyRO7OJm8UAJglGTwIrA5tI+lNDLijo06S0rNdwj5FngN1woWctSW8V7S1tW7VtZ8ezaqKNmS0GXIunzIySNLHzriToD5jZnPig4j482nFJ4DxcNB8haWKpzVoSK2e1y2rP1gq/NRcwBtgLWEfShMZcVRA4Sbx8Efg1cKykj7r5lIIAmNU2XwB+B9wm6ZbS+pOAnYBbJI3phlPsl4TAEzSUNDs2SK2Uf66wX70izy3FwCKFUu+GDxQ2l3RfZ11HEDSCUjvPRR7RnEL7c0n7pG2qDpLT+mHAxrhXyraSHmjc2Qd9lSQSnok/by9OaS4b4GbKw4E1JL2dD5hL+xepWzNaGzSnNrsH7pXylZjhC9qDmS0BrAo8UnjmmNkWeLuqKPK0crz82Zx/nhfYEY+y3DKesUGjydIOr5F0WL1CZBA0mvQ8fAG4GLhE0gdp+WA8ELeI3l1G0kvpc81+bNA5hMATNIzMTHY34Na23tD1ijySfp/y6nfC/UZ2kHRX511JEHSM1D6L9JYWL7caIk/h6VBU1/qtpKmt/M52wC3AVpLubsClBP2EFBExKWubA/AKhRfhKS01RZ7sOIOLdlvcBzDLw+dzwHZ4pMXO0WaDziS12ZG4yDOTOkQeM/s6cI+k9yusWw3YDDgU2E/SvY07+yCYzVPqMTwyfVKRJtu9Zxf0Z5KIcyueknWIKhTxqCe6N2gMIfAEDSGJOy8BPwVOLlTddhynNZHnXeAQvBLRjwlxJ+iBmNlckj7JvreYgash8hTt/BPcaLliCfXsOOsBQ2JWOehMivaZInuKSJ5WRZ4kOP4cj3h4Mn8PmNm2wC7AingKYvhEBZ1GloJViDyX0kokj7mB7aPAQrjZ8gPA68B8eBs+GDdXPkLS37vuaoL+SEqTfRUXdz5Nfx/HffU+qTcaLQgaQRK8H8Tf39dWWF/u5+4CzA8sD3wE/Al4vJKYHnScEHiCTiebcbgKOCkf2LbzeK2JPFNwgWfrEHeCnkCa2dgQn+ldAlgKuA24T9I1aZsWMxudJfIEQSPIPUuoU+QxsznwvPwNcMPF+/A2vRCwLfAHYIykf3bhpQT9hHaKPPPhJdS3xitxfgK8AywHnA38WNKrXXkdQf8j+ZpMwKthfgZvizPwvu4TwAZtSTkMgs7GzE4BTgSWlPRaJV++1BfeA7cO2AOP4C2i02cAtwOnSHq8yy+gjxMCT9CpZOJOYZb5QFped85l8tKZO1d1q4g8iwJFycgwkw16BGY2Nx65tj4+kJ2BCzTF3zMknVhl33pEnoNwkWd6uldWAgZIeqyhFxb0e9op8gwDLsc7d+Azdz/BDcT/rKhAFLSD1AatNS+Sdoo88wOXAd9Iiw4Bfi/pjU6/kCAoYc0G9+cD5+LGylsRIk/QgzCzk3Gz+UMlXZKW5SLP/MA1eGr3gmm3SXj7nYGP32YArwC7SXq0gefa73x/mlrfJAjqo5QrPAw4xcy+BG0uebs8cLp5uXPS/jMz487L8Zfex2l1iDtBjyDNuv0T2AIPrX4RL9cL3jkD+L6ZHVBp/9TOm9LnnwGn46mOTTSXUL8c2MG8ksuOeF5+dO6CTsXMBqVBNFmbzD2k/oJHqL0LPGpmC2Wi4yzkFbG+g3f0wN8Rx0q6O8SdoK1k/QBlYriV1xfLU5sdmKIe78PFmoHAY2Y2rEqbfR9vs79Ni8aFuBN0ISsC5+CmtZ9I2ga4C2+3RdT6l4G/VGvDQdAI8mct8Gz6u4uZbQ+Q2uLSZrYr8BAwGk/LKsaAc9Dcn/0Ab89LAyebV0FuCMUY1MyWM7MNS9fRJ4kInqBTyMJJ58QHtAPwm/ivwLfaEoKffETOAf4N/EDS6xW2GQ38Hhgp6cEOX0AQdBBrNhW/GfilpD+al39eHq9qNS8ebj0Mr0x0WLVZhToieT7E2/8uwE6S7mz4BQZ9jhRtNgpvTxOAyZJebMP+A/BItYvxTtyIGpE8c+PRExtLWqKzriHoP6RJpHuB/+Izwc8D4/HIxn+mvwNx3xxJmlThGE14usBladHqraRr/QIXJJ9rxDUFQS3yVG4zu4OI5Am6gTTGW1HSX7JlGwJ/xNvh28AjuGizKd7fnRuYio8FB+KTOw/j/YW38D7DYTRH9+4q6cYGXsPGwPdxo/wDJF3dqN/qCYTAE3SYNLB9BZ+hHYwPOufDB6NN+A19UBtFngPwyhfHAD9Js8DFup3xalk7hudO0BNI98B4PLrmbEn/S8uLPORVcMPO+dMuTwBrAU2qUgmjJPLsg7+YlsFfmEPSZptLuq8hFxX0adLg9VU8Kkw0h06/AkzGU6mexzts43HBfSLwH0kTzGxuSR+nY62Dp2sthKdr/a/aYMPMlpP0f42+vqBvYV6O93VcIK/EjPTfVLy9zkh/X8cHxE/hg5D/4u1+FHAK8CY+QJ4UA+Sgu0gRuStUSlMppb3cCWxJiDxBF5HEnVeBH0k6obTuUuDbtPTWAW+fA7PlD+PG9Y9Leqd0jGuAfYA7gB2AGa2l3tZ53kUEZxOwGLAvcCNwJPAzSeM6+hs9mRB4gg6RZnD/jt+8h+ICz7m4IttmkSePaDCzW4E10n5/yELqfgH8KqIWgp5AJu5cgZvFfZSWz3q5pNSrkbih3BDgFknbVzleLuzkn/fEI3m+kDbdWFF5KGgHSdyZgM+wTcY7YjPx53dhWl+N6bjp7Ke46PM2HjmxON45ex5YX9K7pYFJv8uBDzqHUvp3zky8PQ6iuR0Xvg4DqM4bwOdwEXMeWpafjgFy0KVk0b/7SvpNlW3yZ2kRyTMDvwcGESJP0ACyZ+91wA8lvZmW55FlvwB2z3bLxZ6ngfslHVnh2IMkTTOzz+PP4CclbdpJ552PJYfh74MNJd1ubvw8ra/3R0LgCTqMmX0OD+1/L30fhleb2J32iTyFN9TewNX4gGELSa8V6ztD3Q2CjpKlJl4JnKqsDHRpuwH4DMIDwOeBsZKONrMVgeG4Ad2/gZclTS515nKR5zLcZDl8p4J2UUqnhdqCTpFuW0RHDEnLipDralQUeYKgrWTtdRgeTTYV9yNZLy2bB49Cq4Tw9joIF4KEt+fpeFuenv4bipdH3yhEnqArMTeifQXv654i6cMa2+aViZ7E078LkbPHRvKEuN87KQnrG0p6qMYE5HnA2sAq+ITPNDzT4i5JD6dtqkX1DsNF9xfxlO8pndVezGwv4Kt4xc5FJJ1rpQq2fZUQeIJOpbiBzWwoHsnTLpEnHWsOPCxwOPAMsJ6kTxt28kHQBrKX322Sdq2nE2Nm43CBZ2dg1/TfAvhsx3jctG7vNDDOX54DcbO6nwOjJd3bmKsK+jJZtNk44LN4xM7PgQeBVYGV8cHu54Av4VERi9BsjFirMMMUmsWgIcBzeKfw3ejgB+0hE3cuwgcABwLvA4dLejINjhfC23ITXlFwKfyZugreludO66uZahapBAD/wCMjZ/PuCYLOJutDPAl8TdJbdewzN95vuBx/xn6R6ulag6qlgNc4/lBgNUmPtOliKh8rnvu9lPTsHY/7882Jp7KuLmlKjQnIYfiz90NgriLap7xdhd9aHTfAP1XS2E68hr3wKl/LAH/Dx5/HJpG0z7fNEHiCTidLSZkDN0tus8iTBrSLA//CByFNwJqKUtBBD6A0s/ExsKmkR2vNmpnZ0niUzn/xQcpK2epin4H4/bF14W+S9l0FT4UcHb5TQXswsznxtncpcDw+SDgGb7+HS3q8tP1QXLBZCo+UmAMXfYbgg+ovpeXL4s/3JjyaIuefeKewR8wkB72L5M0wETgmRdYcCByOG3QeL+nvabtag4eBwBL4s3pJvL0OAZbDhcwBaVnR51hZ0r8aemFBv6fUh/gAj8y9qVZ0enomHwhciPehr6zhybOhpE+TtcEAYG55dbha5zQMFzmPk3RbJ1zj/Hgq7yg88m4hPPruoyKVPeh5pL7Cc7jociH+3DwPb18jyhFimR1BnhZVV6ZF8p4aA+wFrCNpQgfOOxebBgMnAi/gbfAiYFE8u+SENEbt0yJPCDxBQ+gkkWdz/GUA4TcS9BBKKS6T8VniyXja1N/KL7YsrHof3Ij8TOD/8DSWnfFUg7VpzqefBIyRdHHphbmKpKe76jqDvod56dL7JL2dvu+Biz0fACcqVSSsJ7w/69QV6THCRZ8F8aifLwE3hCgfdAQzmycfDJoXYDgCn1E+RtLjaRALeOms9H2QpKl1/sZwUuXPeqIogqAjVOlDfIyLPL+ulD5SEncOl3RRtq5ada31UsTF3sAFwLKFlUKV478AXC/puLSs3QNgM9sC2BqPDi3OZ1Pcs2094CylYhRBz8LMlsAjeh9RMkRO/56XUEXkaeV41dK65gV2xP0rt5T0QAfOOT/uGFysPyhbtihe1XkY8B1Jv07L+6zIEwJP0DA6IvKY2SL4C+kbpNzPrjvzIKhMluLyMPAZPKXF8PY8GRci/561/ULcWRWPwPmepLHZ4Hgw7h9xDZ4nXNwbN0raNf1mv8gXDhpHrTZkZrvhIs9HePv8S6Xt0raVZupmpQHUO2sXBK1R7nhbS1PPXOQ5tl4RsRCCUhseAMwst+cgaCRZH+LveETBcvh7fwDuXXIg8Jv8eV1N3DGzIZKmpM+VRJ5H8ZLUFwF7VovKSVFuTwArpt+4TNKLHbjG9YH98ZSblyqs3wzvO13bWlRR0DNIz8uRuMgzkzpEHjP7OnBPpX9jM1sNL1d+KLCfOsl2wMz2A36CR7vfWpx76ocvgZdyfx84QtLdaX2ffP7XyqcPgg6RBrhN8nz2Y4Bf4jdWE/6AWA+4wszyVJXiQTIKF3c2C3En6AmksNUJwGV4R2oM3oES3p6HAg+a2Vqp7Q9JL5XV8JfKicryi9NLZWqaUdsPn10onsmLFduFuBN0lCozwk1p3XXAWXh61ZlmtkG2TQvfkqITlHeGlHk8ZLNl1fxOgqAuyh3u9Cwt2uxV+ATQZ4FzzGxEvcfM2vCMSu05CBpFinacgKfJboFH876Mv/dn4JM9VwJfT/3gQtz5FrOLO00pOmcggKRt8Ij3gbhYNBGvQvszYJdq4k7adzpwL17J6DDgQDNbuJ3XuBme/nuKpJeKe9ac4v4tfutoM1uwPb8TdB2przoDuB/4Lt6+HjOzYXLP1dkKLpjZILzq67Nm9h0z+6KZzWNmS5jZ4cBv8KqbO3eiuPN1XNwBjxSbde5J5HkVWAf3dT3fzLaEFlGfxXH6hDbSJy4i6LlkIs9k4FjgV3g6QCHyrIuLPCvDrBtrR+BHwFaS7u+eMw+ClsgNvg8Azk/jhNvxDtpjzC7yrJs6X6vjws1Jks6FWR2zWQON9AJ6Dy+hXkQ/vFKs68JLDPoRxbM5fa4o8rR34BsD5qARlNpsu0SeIOguJH0CHIxX0Zwh6WfA6cBLtBR5rgB2MvewOQBv50eUxJ2Z6ZizBtiStgbuwEWeYelnN1YN375CSAJuw8Wg64EjgUNTJH3dmFcF3RO4WNLLpfNU6f79My4Y7FJJIAh6DlmUYyHyHEwrIk+a+NkIeByP+nkKj1y7BxgL3ATsquSjVol29H9vxU319wW2S6LOLEE/RRu9CqyFR+CfVxJ5RqbPfSIKOVK0grqoFYZX5/5FyspQPF1rD2ZP1zoSdzu/Gtih1kspCLqSVlJcRuEzFSNoTteahM+6XQX8UNLZadvZQkGze2M08Hu8GtHmEbkWdAXWMne97nStIOguSm22XelaQdCVtNKH2BvvQyxDc7rWp8DdeJRDOXJntgGoNVewNbwq18q4L+Cf6jy/hYBv4pVrd8cjjM7GxZq6fKnMfU4WkPRMEnKUBs4LSXo76+vkKer7SrqinuMHjSG1GWtN2Mj+3Yp0rUtpxZMniZSX4RkZAIcAv5f0RidfQ4vfNrPtJd1cYbsiXWtJvLLWO3ifZzlcuFpbVXyqehsRwRNUxczmTh1+yjdtW6kQyfNLWkbyrAf8Go/wCXEn6FFU6pgVswvyPN/TaY7kmYFXHPoZcJWks81sQD6bUDp28VKdN/39JxBmykGXUE8kTxD0JCpE8ozFI3nOjUieoCdSpQ9RtOFyJM9M3IB5B+DnmbhTdRCexJ1huLHxcnhlz3rFHZMb7/8GT9X6Hh6pcRxwSD2RPOlatsHTX5BU+FstC9xhZr/HIyfyVN+pwBfNbPG+khbTvLEeVAAAIABJREFUmyiit4roqrTMyuuL5enfc2Bqy/fhYs1AakfyvA98B/htWjSuNXHHzA40s01rrG/KPg9LvzO9tH458yISLVBzutYEvD3Oi0f+HAVsIem9vhI5HzdUUJF0Yz8F/NLM/s/MTip39vObvx5KIs9xNHvyFC+sJYCRIe4EvYEibDV9zkWeAgO+ae7JMwMXfyqSZr52SF8PVJQQDbqQekWevtLxCXo/RSRA+vwTXORZhEzkqdReYyAZdDVm1lS0xbxNlp67uciTt9sdzGxXMxtcR+rrSDzVanu1oSJRdtw5ge2AZ3CR5x7qFHmSQPApXiE0ZxiwCjAan8gFWtyHb+FRy5FO0oWY2XzA38zsdjO7ycxOT9GQu5jZyma2FLC0mQ0zszkyUW56+jsTbx/fwtvr462IPAfgNgRTWjmvE3BfnvuqrM+jN78L7J0iwfLfm4kbi89lZnOU3wNq6ckzDngdWF/SK9UmYnsjkaIVVMXMLgP2wX1FhN/EFwDPpQ5VsV2bKqdkIXKDgPPxG38IbQgnDYKeQp52ZS3TtcBDrWerrlXavwkPjb4CNxUP36mgW7CWlYrqStfK238QdDVp4Ds1fd4fT/V+C0/XejQtbwI+J+k/3XemQX/FzOaS++8U31v0A0qD1krpWp/gA+mKJdSz46wHDGmLuFPhGFsAUyT9ybwa1gnA5tRI10oD6IF4isstkl7O1g3Bi0hsgqeavVHa91C8aleHsgSC+jEvT/46zT5NZWak/6bivjkz0t/X8SptT+Emxv/FU/pGAafgabIbSJpUKV2rjvM6DB9jrifpkQrr87723sBPgeWUKrWZzVa9dh/g/iTkVPq9HfGskZUlvWh9rGJtCDxBVazZEwT8Rs9V0gfxG+N+Sa9k+7RV7NkD+Dkh7gS9mBoiT6US6vkgeiAeuXMtnpp4d3ecf9B/KQs0VUSej4HjJf3FvJrc5sDnJV3YLScd9FuyVIFVJT2VL0ufc5HnUEn/TMvOAr4s6bVuO/mgX5AiCjbES0AvASyFR9bcJ+matE2LwWRniTydcO5DgTWUPADN7CvAibQi8qRtRwMfSnqw2nul9H5ZE0/zOULSu426pqCZFLnzGh6tlTMTF28G4f3VgenzDLz9VeMN4HP4RNA8eBT7hm0VecwNup9JX38DnCbpmSrbHoEHBwDciBuL3y3pv6Xt1sG9pC6Q9HGF4zQBC0l6q6+JOxACT9AKZnY9bo41Hb/hp+I3+4C07CPgNOB5SXdm+7V6s5jZN3BD5Z0iLSvo7bRF5EnbDAK2x716wncq6HKywfLOePTYQWl5JePl9/BSvQvhVTG2C0Ey6Eqy9lr0HdaS9Gxal7fZ/fGB41R8tnkvYFQ8Y4NGY2Zz45EF6+PPymKAXPw9Q9KJVfatR+Q5CBd5ilSYlYAB6iRz8SRO7Y/36e9Ly9YDxtB6JM/awLKSfpW+532isuAzCE/bekFSeA52AWa2ADABj9z5AH8+3oWnzg3DBZq5quwuYBou+kxP34tx4JD0dzqe8fEosFE7RJ6jgKIS2y9wkefl0jYb4VE+ZwNfxKsur4SLVmcBT0l6ONt+NC6sflJqj7nQ2KbAhN5CCDxBTczsWPymuRVYFZ+NKJiC39jFi+vnuOv/zZImZseoVDlobuABYEwuDAVBb6b00qgm8mwg6TEz2wXvCG4fA4+gq8kGy18HbsAFmzuy9flgYyfgXPz53wRsFeJO0JVUaK9flXRbaZv8+XsMPggAr0pY0dMhCDqLNIB+Apgf+DdeKXYxfNBb9JcBviU3B690jHpEnm8Cd+KmxtcDqxfRbBWOVxZWWk2pTVE8G0v6Y7asXpHnCOB9SdeWr6e03XzArsD1kj6odT5Bx8nEnYvwqJsDcQ/UwyU9aV7taiHcrL4JF02WAhbAPZSG4iXIP0tLj6icIhAA4B94G5pUx7nlwv1ieHs/CI9sP71IwUrbzg8MlfRmtuxAPFJnx7Toajyq5w/AzsAiSkbl/YkQeIKaJDX/SdyI6tt4qOa6+EO+8OWZjN/8pM8vAj8CnpD0eOl4A4C5JH1opbzkIOiNtJLiUknk+QTvIB2Pm8mFuBN0KaXB8nW4yHhbuTNuzTntTbjIvzWRTht0MaX2ej0wulJ7zbYfgnv7XYQPMv7cxacc9DPMbDg+gL4Z+KWkP5rZYsDywE14tZ6JeKTExZIOqya21CHyfIjbJ+yCR8BXnCQ1sznwCI37gXuLtKu0rmKUvZmXzMYHxnfn4ks1kadCH+hw4ANJ15YjONL/EwO+CvwxH7wHjcPMrsHb3zEpsuZA4HA8lfX4LLK8ajRLihhbAk/vWhJYFhctl8PTtAakZQOBxXFvm3+14RwH4m36VTyKbA98EvSMcjtJ7XRAqW1thvsBfQOv5vYXXOwZARzVmrDZ1wiBJ6iKNbvcnwx8HdgAV3wND3neBL8ZCzVXNOdwTsLD+cbiZfGKUM898BfDiErqfxD0JmqkuJTTtb4PrIFHuxVeVltKuqc7zjvov7RjsDwY2Bu4DG+zYQIedBltba9pn13w1NdREWkWNJok7owHLgfOlvS/tLzwnVkFj1ifP+3yBF6iuUnStCrHzEWeffA+xDJ4Wk0RCVQ1Ms28fPQ/gc/j/fHBwHnAnfUInub+OGvh99GnWX9mXVzk2QLvy1+jZpPbEUWqWIr+fwuPBPktHg0yGVgHF51eVPhhdSlmNo+yCq3mVbOOwM2Rj5H0eBJOgBaVYgcpGdnX8RvDcaGnqS1jvGwyaT3gOVxEOhPYnSoiT7ZvWWBcGPguHuH25bR4tLzabb8hBJ6gVcxsdTyncr8i7DItHwSsgKvAI/AwvkpRPe8DD+FRQGOAXcuh1UHQ26gjxSUXebbGhdI10uqYVQ7ahZktCHxD0qXt2Lc9g+UvAY/jKTExWA7aRa02VmOfNrfXtN9PgJvyFJMgaASZuHMFcEoxgM7abjFwHYmXiR6CV5ravsrxcmEn/7wnHsnzhbRp1T5EimD7N25WC97veBv3vBkI3AKcCvxH0v+ycywPlNcEnpX0aSkyeV3gJDyS51JcBDoFeEDS+dn+g4GN8ajlpfCB++sqmeEGjaWVKPNc5DlWdXo5FUJQauMDgJmFINSRSBkz+zKwKfBjPAroe7gnT02RJ9u/EFWLNv1NYCQucp7YnwILQuAJapK9pK4CvgJsqiz3MW2zDHAlsBou5IzKVs+g2YQLImoh6AO0IcUlF3l+hBt/RopL0C6yNIBrgcMqhdjX2Lddg+W074ptCbUOApgVRbCTpF+k73WLPB1pr0HQFVizr8mVwKmq4iWTBsCL4VE8nwfGSjravHLQcGBBXJB5WdLkPK2pJPJchnuT1OxDmNnmeNTFHUlkORg31X0Sj6pfH48EehA4HfhbNuBvwsftMrMt8SpDxf2b92fWwwWnrfBJ3Z9L2qe8XdAzKbWrdok8DTy31XAfpwlmtjJwNG0QedIx8rY6HM84eVjSG/2lfTa1vknQn8lugnF4nuUy0KzemtmiuPnml/G0q9H4A/96/MU3gGZxZ+MQd4LeToWBR0VxB2bNbgw0T+H6JrBFiDtBezA3pRyPhy3/oCvEnTQwIcSdoJ3sDFxhZvsBqNnPqSYdFCOrGYAGQaeRnscTgNskHYX74lRE0gxJE/AB9JvAb83sIlxgeQD36LkNuMnMhssrZDWlfWemPsROeKpsPX2I+5O4M0CeWvNjXGCaB49mWAc3ad4an5S9zcy+U/xe0e9X8gc0s5XS9yJlB3mlop/j4s6lmbjT1B8Gz72d/FksN/y+ADdQPsfMRnTrybkh+WoA8lLp5+JVtfYFvp+CCmpSaqvvpsUnpnuiX7TPiOAJ6sbMHgY+krRV+r4oXi53Q2BtSS+a2WBJU81d+BfHX1wrE1ELQQ/Buj7FZUU8xTGqZQXtIg0mXsPz0jeR9Kc8RLqVfSMSIugW0szwFcDHwNFpINGakWe016BHU3oef4xHtj9qNUpCm9nSeJTOf3HbgpWy1cU+A4GHga0lfZztuwrwd/xeaFMfIrufhgLHAM9IujmtG4VH8xyTNr8LN4n+taT30zZLAvNLerLCsf+Ip1x9M32vaNwc9Fy6OpKn3ugZc3PxN4qggNSPPpY2RvKUjvkt4FfqJ8V9IoInaJVsxu13wEpmtmYSdy7FxZ01k7hTzBaAp2Z9EY/62TTEnaAnYM358l8sohPasG+7Bh4p+mGNEHeC9lAaTADcYW5mWfid1do3BstBl2JmQ8xspJldT3PZ2rmAS8xsX5g1ezxb2432GvR0zNOyiufxZLx09J/NbO088ibbvuhnbIgLOD/H06S+gkdN/C0tN7wwyUrAPmnfQsR/GlirPX2IIpJB0mQ8EmJF83LUSLpV0nF4BP6ZeJ/9x8ATZrazmS2XIo+2NbMVSte1BfBoiDu9mwqRPGPxSJ5zOyuSx8yWShOrdU1IpY/XAe9k5/kv4ByqRPLUity0BJ5R8uVq2/U1IoInqBszWwQ3SfsN3mHbEn/pvFR+uKeXwa3A4ZJu75YTDoKMbKB8FnB5FrZZz77tTnGJTk/QXjJBck5cNJ+Cl9mdBGwo6bFKbTCP7onBctBVpGfsb4BF8WftX9PnwfjM66fAD5UZsab9or0GPZ7sefww8Bk8Ot3wyfLJuA3B363Z4LUwfF0Vj8D5nqSxWX9iMN6XvgYvGz4zHetGSbum3+yUPkQpkuco4HlJN2XrB+Gl3H+A+5V8CXgDN7l9HRgo6e5K5xP9nN5PHlljZvsDR+JV0I5J/YzZIm/qeTab2XG4eHgf8DtJP87aYq2It3mB/YA7JD2fLc8jea7Fva9eqXKM8rh0Obx0++9q/9/oG4TAE9RF9qI6BTgRD+HbKIvcKT/whwPztTWELggagUWKS9DLSDNeL+EzanPgnRqAiTSLPBul1ICK5t5m9jW8wlu02aChpMiGl/GZ119LerC0fitgf+ClFDVQ6Ri7A1cDO0d7DXoSZjYnnl51KXA8XoL5RLxCVSWRZ4ikKeaGsX8FTpJ0bjpWi75HunduBdZNPzdO0gYNuIayyPNcMdi1lsbOiwG74oP8hdN1PQJsVkTf1ZNmE/QuLFlspM+5yHOspEfT8ibgc5L+U8fxfoDfK1PxNMT5gb/gflNjS+JLJQFpadwv6rfAdDUbgecizy/TsZ5MkWkfSLqrJFh9Tm6uvCywklKKYl8nBJ6gTaSwzD8Cu0m6oXiJlbaJh3/QY7DZU1xqRj+U9g1xJ+hyUifqTuBVvFrWxExch5YiT8W2nNrsDcB2csPNaLNBQ0izrc/ifjvnSpqUlg/GO+aFx8PwInKyNKAcBCyEP6d3lHRztNegp2FmuwL3SXo7ff8q8H1gBC1Fnk0l/dXMVscLlOTiTjUx/vt4qfEm4JeS9mpEX7qCyPNvSb9N68oRD8sDy+KpMYvjEfnXdOb5BN1H1hZWlfRUvix9zkWeQyX9My07C/iypNdqHHtn4NvAN4AdcPHyceAwvFjPBFwsvVuegljsV74/Rkq6v7wuiTxHA3sBN+Lpjavh/lVvZPtfnn57Mzz18Xf9JfAgBJ6gzZjZ7/EbdENVKQsZBD0BixSXoJeSOuAGTM46XKfipWmhFZHHzDYB5ow2GzQSMxsI/AqvlnmwpNcrbDOr9HIrxypmWqO9Bj2GWilI5kbFJ9BS5JkEfAu4Ck9JPDttWzXNxcxGA7/H+yibS3qogddTS+SZLbI5RZPug5eZfrhR5xV0HVkb+AYeNbmWpGfTurwfsT9wCB6F8xQuqIxSDT+o9LxfVdIT2bID8Ciem/A2ty1+z/wPOAN4oBCZSsfaEfcEuiydby5ArQj8ENgJ9+tZT24ZUtxTg4FDga8B/wJOk/RyO/+X9TpC4AnqJnsg7AOcDewk6aHUwRtUzNoFQU/AIsUl6GWkDglqNqsvlufRDq2JPEVp8yKcOdps0DBSOsezeFWTsyqsL9K7i2fqAbh3yQrAu8AzuD/DY3KT2qq+DEHQUygNNHORB9zMVXj58EPLz+Qqx9sL9xR5DI8A+qiOcxiYjjs9W1aXH05rIk92bcX9G/dlH6EUmX4D8FVJt5W2GZD1IY7Bx3zg4uN9bfitQZKmpc9H4JWxbkzfjwX2BpbHhc1f4yldH0uamB1jC+DJLHIub58/o9kP9tUKUWhz4N5WQyV9WO959wVC4AnajJkNwdXQByXtb543vzZwXIg8QU/AIsUl6GUkQfIG4ABVMA0sdbhykWcyMJSSyFNPJz8IOoqZ7Yb7IBTeZnmHPp9JPRjYGBhV4TBPAbfjhplT4lkb9AZqiDxF5axJ+H3x91pt2rwq7SXAaGBEHvlQZftjcOPyUfjg9RbgiVpRFbXOv5LIU+06g95NSdxpNTI9jfcOAC7C/aX+3N7fTJ8PBaZIuiJ9Xxr4Qjr+srix99+AsyT9I22zBjCs/Nvp3XMssIOkVyqIO/263UaZ9KBNJBV/Cl5u8Wtm9ju8bN0fQ9wJegrpRTUaD8+clJadBJyeNhmGizxz4CVORygrF5l4B5/ZCHEnaCjmJpvj8UiINyttU8yips9jaJ5RG4KLPHlbnlFqy0HQKIoO9F7mJp3TsnY607z65q3AyTSLO0VfYXraf1XgQOCHSSCKZ23Q4ylSRtLnW/H+xWN4OvgM/Jn8oJmtVaF/AcyajNoG769sVoe4sy8wVdLpuAfJ9bg4c4eZXWZmCxfnVPxt7fzlJdTPB1Yw904pfmv9Yrt6/n8EPZu2ijuJHfC2sVV7xB2Y7T65CBhkZgel7y9L+iPuk7M98HT6zXFmdqmZbSc3eF7fzL5QOvQrwLaVxJ3id9tzvn2FiOAJ2oU1my2Dv5Tu787zCQKIFJeg92HNJuDXA0fXGZo/APdEuCpbXDEqjTq8T4KgvZjZCngJ6Lfxcs9jJU0290cYiUfufAEXc5ponlicgafQDgGWTMuexI1c/9zfZ1+D3kONSJ5K1bXySMyB+GD2WjwK4e5WfucYYJKkS/LfNbO18HLmo4GHgFNxT5Pp9dxH2XGGpHP5ANgcuEvSwe34XxI0kPb0Sdsp7mBmPwFuSiJMhyjdJ9/FhcorK2z3DWALYN+06Aq8j3Rv+R5K20cfvQIxwxe0GfNykV9MXzcKcSfoCaQUlzuARcvrUkenEG7G0DKSZ7boh9K+8eIIGoK1rPD2ET4Ibm2foXhk2lV42enCW6piVBrxng86CTMbngaZOVNwo8ylcf+EJ8zsLnygeTYu7kxL2zbhos6VwJrASsB6eDsGr4IyGmL2Neg91IjkEcn/g+ZInkLcGYSLOz/DK8e1Ju7MBwzPxZ3s9/+OVzs6GY+GOw+PsB+Sn1sr5z84Red/iBtE/ynEnZ6DmQ0zsz1hVmRk3e/19oo76bf27wxxJx0rv08uAQab2YHZeRYTtDdK+iY+QXAT/k44DbjQzOaJPnp9RMcvaA+b4i+QLSX9pbtPJggixSXobZjZ/DSLO+Ad9GPMbJ4a+wzFU1nOB46UdI2k8/AqFDC7yLNGastDzWz3NGsWBG0mPWMnAEsXz1EAuV/U8enrXLhh5ubAvHgEA8AgYCDwF2BP4HuSnpQ0XdKb8tLL+6RttzGzBeL5G/Qm0uC1mESqJfIURsw7AT/HI3fq8c75DLAOzDKubRGZme7DS4AxeEnzM4H92yDyFFHPo4Gr0wC7iBYNup+dgSvMbD+oX+TpiLjTWptpD7VEHklTi/aWzvtB3IR5ZbwU+kx8QiCog3iBBu3hdWAbSfd094kEQZrZmgBcB5wozyevSJamNQB4sTgE3vmq6MnTiJdc0L8xs+F4m52Jp1QVHa0fAEea2VwV9inEnQvxNJYLswHFiVT2l3rQzNbGO0W/wHPWg6BNZM/YX+NGyMVztOio/xo4qMKuQ9LfvwKnSNpI0iOS3s+OXfRDb8IF+jmAGTErG/QGSpE0M+oQeR4wsxPwCMzt6xR3ClYysxWUTMzLSHoXNzz/IS62Hg3skFLUW42IM7NNgJ9KOiB9D7P+nsMQvP1cYF6JsFWRpyPiTjp+1TZTrV9cT3+5FZFnVsXFtPmU1K53x1Mf2+UD1B8JD54gCHotpRSXscBJysorVtlnKPBtPApif2AB4Jy0uqInT3Rygs4iRUK8ire/cbiosyYt/UlOAc6X9HHaZw68kkUh7lyUljfh7/FK1bWKtlx05LZt42AiCIpIs1fxZ+zBkn6cls82SDCz0fhM84p41M5HwG+BOyU9l7apWm7ZzJ4A5scrCb3boEsKgk4hG0DvjHtRHpQvT59HAd8H1sB9pwan3besd5I0DYYH476X9wIX1OrnpHfMnvi75VVc6LmvNZHHWlbAi35PN2PuifQV/N0/H14OXHgq90GSfpq2m81nqaPiTh3ntnA6l4VxQf7ZbF1dbad0n7Tw5LGSV1X+zqh0vcHshMATBEGvpDTwKDgZ7/xUNKotRUEcKenCtPw0vBMGLUWejSQ9mvbbCZg/zTgEQbswsyNwU9kfSvrAzNbDU17XwMvrFjNgp+BphMI7eD+iJO4UHTWrXEK9mDUeQCph3UWXGPQRMjHyRbxdrogPTO8rtb+8o54vH5pHVNbqmJvZl4C7gOskHdfQCwuCDlIaQN8AbCfpjvL69HlrvG+yRlrd3nLTF+ORDDtLur/WYD3du3vghstPpH3+F8JN7yBNXv4G95R8DY+CXBQX+vYEPsX7EOeX9isiYxop7qyJR2x+ivumrYN7ro2T2x+05Vh1GS8HbScEniAIeh0pxeUVfBA7EA9fzaMfzpP0SWmfcorLRVUGxtAs8kzEPafmxGfORkm6vZHXFvR9zGxYPgNrZl8BzsUHAHkkz3l4GzyJKuJOdoxqleJC3AnaTHpe/hV4HB8kzg2ciEfobCPprirtsOrMa43fGgYcARwDrCXphc69miDoPEriznV4qtVsA+jS4PVHwCHU8TxOgtDdKlXyNC8k8WDabANJ77cSETccN+QfA1wmKTzYegFJnHsZb1u/lnvR5Ou3wqPPX6omhpvZ7sDVuLDXmeLO2sA2eCrf+LRsLTyCfj28fY4BHlMNu4TSMfP75DvAYHkK+pHAO5J+0dHz7o+EwBMEQa8iUlyC3ko5gqEU7VBN5AE4TtK5abtWZ2DNbAs8nL9dM8VBYG6kPAJ4XtIHadkKeCRCTZGnyvGKQXETNFc+SZGYOwKX4lEQ9zbmioKg47Q19SXdR9vj5ce3b619p8H7XpJ2q7BuAHAw3o8ZJ2mD4jdqiDxL432l7YDNy2JB0LMws3lxL7IrgHMlTUrLBwPTs+fmcKU01tLkziBgITzqZ0dJN3eiuLME8D3gNEmvp2UD5L45iwOHA/sB7+L98lskfVxPSlV2X80J/AlYCk/zXU3Syx099/5ImCwHQdDb2Bv4CZ6KdTc+6/soLsQUL5GTgKPMbI40E70/FcQdSTPV0hgxL6E+B54z3wRsGuJO0FHKnZy80yVpHC3bMtnfGWY2d+oEVRV3zNkX+B2wdYg7QXtIz8bpkv5WiDsAkv6NP1tvAu4wsy3VutHnt8xsO1J/Mz1zi0HKOniEwfG4gB7iTtBQzGxBM2tX+e+2ijuJ5XGD+53qbN+bA/Ok32txX6Vn/8/wyIyvmNmD6ZymW1bZrrTPy3iqbxOweh2/H3QT6d/wSuAfeCWzScU6SVOVFf5Q5lGWi3uSpiXxZbHOFHcS8wOTJL2ehKTcFPk1PNLzSFyYORf4VhKi6qnipvT3U9wg/GNgFUkvW1RyaxcRwRMEQa8jUlyC3kw5CqcUopy35bzTfhpwjkqphxWOfQ9wkbyKSxB0Oma2PB4l2Wokj5mdh3f6rwLuxv1AFscHst8GngGOkvRYV51/0D+x5uqF1wKHtRYJWdq3I+WmV5T0rzp+Y2e8Ut1BquBDkqVqDQeuAUYBD0oamdZXje40s5Nww96tO3HAH3QiZrYYHr1zhqSzKqwvomWKtngAXkJ8BTxq5hngPjw9anqtyK42nlfxe3vgQuUONbYZhkesnYYXMDkf7498WOdvjcYF0VUlvVJPxHJQmRB4giDoNUSKS9BbSLNOC+OGiPMA7wMvSro5ra8VVr8e3pbXpKXx8qm4v9THNX53jnzmLwgaQRtFnuOBM9LXKcA7uFnoecAvJb3VRacd9FOsueLmWcDlakOVtvaKO20ZnGbiznPArpKernVM8ypGNwAb4b4n26lCZa3s3HfEfa42jgFzz8TMdsPL3G8i6U/WsqpZIe4NxtP0NsYFvjJPAbcDp0qa0pkRPOZeO3fgFeOerLHdMGAH/P0wN57WdV1r/ZIU5bMuMF7SGyHudIyKIX1BEAQ9kdZSXMzsGFqKPDPT3xlmNjfwSWspLsA+wMVEikvQTlIH51RgW+AL2aoPzexrknZPM2wVOzCSHs7aci7yjEnHnyXypDY7DzAxhWeHuBO0mSRIzizC6cvP2jKSnk9RAeDpWlVFHklnmdlEPE12CJ5mMlbS+424liDIycSdOYGHJL1bpIy01s47ErlTz+DUzA7CKxEdCEzGozSfTusq3Usz0nvjv2a2Cy7ybAzcb2Z7SHqx/BN46vrtwOdjwNyjKdriXmb2V0lTi4mgJO4sgj8718XTmMCrvc6Bl04fAKyKV9saYGYnFQJRJ/EW8AGwrZk9X62vIWmimf0+fT0d77f8D/hDrXdLWv4wzGr70VY7QHjwBEHQ6yjn5GadtdzHZDotU7WOoWVJ9dlIL5jd8Bm08NwJ2kwaTDwMHIaLO1PTqml4p2xXM7sLag8AJD2Mt9l/4F5QRadoDHC0mQ01z9nfFfg9nh8fBG3CzIakj0OLjncdg97Babvn8Vna3+Iiz1aq4skjT489Kn0dH+JO0BWUxB3wdjoitfGaviAdEXfqPLelgO/j4s5fgf0k/SytG5hFJ89rZkubGybPem9I+i/wjXRuawE3m9nG5ka1pAiQIurjSFwMCHouTwCf4FFZR5vZ0DQRtKI/HsMtAAAgAElEQVSZHYJHam0ODKXZn6/winwZLz4CsCCwJS4Ezeoft4VyHxtA0qvAzbh9wchq26VtJ+L9khPwaM0xZjZPup9a9dSJNMKOEylaQRD0SCLFJehtmHsjvIyLMr8Fnsbz4w/Fc+Wn4ULNTDxCrFVvpxpt+Sx84DIWL4V6e6deTNDnSYPfJ/DIgWnA8/gM7du4n8NU3OxyfFr/eqUZYTNbEm+PX8fb9d010rWOwn1DwnMnaCjpeTweF3dm4OmBw3ChY0NJj1Vqp3l0T6PEney3tsa9qG6R9JO0bLCkqenzj9Kmm+DeVf8HPAScoOYKS/MB3wH2SNtcg5fYflZexehYYIKkGzvrvIOOkdrmfkrWAWnZUriHzudxoed1XLRZC498HIo/hw3PwHkpbX858E/gM8DWeBES8EIkhajelnPbBI+CPz8TGYsUsc/jIs9ngPUkTWilLz4X3v85DbhBFarDBY0hBJ4gCHoctVJcgDsk7Z62q2UqWJfIU05xacDlBP2AbDBxNXChpPHZusF47vpIvIM2CNhT0q/qPHbelsv+UltKuqcTLiHoR5iX430dH/AWqayVEC76DMDb9yf4YPlv+PP4JeBfeNvcF581Hinpz509GA6CejGzBfG2ORaPcjg2rZpIs8izkaRHy+00i9z5Gp4C1RBxJ/3WYGBtvNrW3yQ9k637BjCXpKvNbFmgKHm+Ih7xcyLwhKQPzKsaDQXOSeuXBV7Ao0n/I+my/No68xqCtmFmC+DCzS+AQ3JxJAmKN1TYbQou8hT8BTgOeL4cDWlme+FG4s/jxtof1NNus3Z/PPCxpEsrbQMcgqfbjsdFnrdaEXkWAX4EbAasK+mF1s4l6Dgh8ARB0KNIs1EPAivhg46peIjntPR9AHCPpC3rOFYtkedMPI3r63gZ9V0kvd2Z1xL0DzJx53Lc3PCjtNyAAWouY/sYHskD3rG7tDRbXHUAYW4ifg7wZbwjD2ECHrSDCmkrZYQPgAdn3wfhz8uB2TIrff40HXMmsKncKDREnqBLSSmCd+KD6MOSJ8gpuCACLUWeipE82UB7O0l3NLIdm/sDror3eR6R9KSZ7Q38r4jMzCIoFsej5XbEI+2OBX6Dv0KKSoxNeCTPDHxw/0laHuJON5M9e38DnCjpjbQ8r6R5IN6XqMRf8f7vDyscu2gjw4C/4ynhq6qOClbWbN69KPA4cK6k80rbFALQADwi5zg8Ynn9aiJPts/yuIH4/pKuae18go4TAk8QBD2GSHEJehuZuHMlcEqlzlTyQphmXtb0Yrwdbqfk82RmC0h6L9su7+zlnzfAw++XIVXa6IprDPoOafZ4PC7WTMMHuT8FRgDDgSXw8HvRcsa4TJEqUDyPB2fHLKKCIpIn6BbMbCjePidnz89TcU8QaEXkSWkqczYqcqfC+RYiz4p41M28kh5K64pBcl4mfX/cv+cVXEx9txhgV3p/hLjT/ZjZ/LjoOCdwsKQfp+WV0gRH4xUKV8TF9Y/wPvGdkp5L29SKmnkC9+UboTorxqVJqEdwcXBNuedOeZuiDQ7E+zz74JFyX5H0dpVrKcSjM4BPJZ1ez/kEHSOqaAVB0COokeIyzsx+QcsUl6HAYvUcV7NXJCpSXI5Pm0SKS9Au0mzceLxtXVhE7pTJUv+exAfCzwKPmNmJuHC5ppm9CTxuZtemwYYpkX5rEF5ydClSdEQjry3oe2SpAecDDwAnAQvg3mYnZ9sMwQWaEbjYsxjeTptwf4jPpvUL0OxvYvhAZFD6uSbgQTOLKLOgS7Bm8++pkiZny4tKRGNSwOQJuLhTiDx/NrNC5BmQjvFA2rdLxEm5V85T+H21Uzqvh9K6wvx8ZnovvGtmV+FmukcBP8AnwWbk25f2DXGnG8mevS/iz8qLzewFSfeVtive+38ws1szwXFoqU1bDXHnS3jbuK4ecSed25K4YLM6XpTkv5W2LcSdJCQeiPdndgMeNrNRhfhUorh/3k/nFXQBIfAEQdDt1JHiMtXMtqJlist82TY1U1ySyHMskeISdBKlNJcPszZb1RcKNw0HF3p+CmxPc3rL53H/koPSYOOvpX23An6NR/480JnXEvQblsY78RdKet/MPsWfiWPMbJikq1IkWTEzfEelg5hX6RmAt9nh+D2wLJ4SMBxPNbH0d2KDrykICs+dG4AD8KiWWaTB6ABJMyqIPJNxj57ZRJ60b5dFnmUizxBgLXOD2k/Lgk0a3L9nZucA2wBfKtZ11bkG9ZOiye4DbsTtAebG0wXvNrNtJN1V9F1L/4Z59asWYk61f+uUnrVD+o2r6zi3LwOX4QLPIukcT5c0pVpfphCW0n21Dy4s7gncY2Z74ib6xcRUfozn8PdG0AVEilYQBN1KpLgEvY1SqHXBtyVdkdaX/RyKsOYz8Mix3+Gd+Dlx0eeLadNiRvl+YG9Jr2fHOByvinJ3464s6OsUz8js+5q4yLMQXjXlmrR8Vsfc3NOjKR8o1/MbZjZvped5EHQmWXTENcCxeaRDabtZKS1mdibuISLcwHYorVTX6iqKaE1J79XYpkh72QM4GVhNNaqDBt1HSmcagRsif5CWrYD/u+0MtBB56jjerJQ9aBYhU79kR+BSvH98bx3HWg74Ax6leQZwkdyzKr9X1sWjNZfExZyH8KjPKdn1nQh8FxelTsBTySZkv7MZHoH/I0kVo4OCziUEniAIuo0sCqJmiku2/Zp49ZZn8eoAh5BSXIA3cXO4Fiku2b6DgM2BW4DNIwoiaA9JkHwF7+gMwKMWCr4j6fK0XSHqFB3x9fG0mBNwA8NCdFwB2BsfbICHM78DbCXpqS65qKDfYS39RqqJPHUPckvHC7+PoEvI+hDXA0e31odI+wzAvUOuyhZX9OQhMy/uDlq7B81sJ9xHcKUQeHoetf79zI2HT6FOkcfMvoVXPryzLLKb2TrAlsAewEHl1K9WznFNXBi6XyW7AjPbCDgQ73dvhJdhHwg8it9DLxfCPzAKL1qyC94X/y0wAfcP2gg4M03EhidbF1CtLGYQBEFDqZXiUmO3copL8XL8PJ7ecjDubbJuhU7ZVvgLJ1JcgnaRZoonABfgnZl7aZmCcpmZHQSzctUHJ3FnVeAe4HhJ55Rm3/4t6Xv4ABv8vbwwsEn2u3modhB0mLyDLekfeDWet4GjzGy/YpuinbbxeCHuBA2n1If4iFIaS5V9huJ+NVfhRsVF+fTCk6dI1xqR2nS3jZOSUDrTzOYws3lK6wqLjdeAK0Pc6ZnUEjIkPY/7oN0E3GFmW7byzF0On6C8zMx2MrOlzWwjMzsNuB3vM+zSFnEn8SRe1WtxM9u4WGhma6ffPFLSRZJ2AhbFI37WAe4GvmVmi6T0x5sl7YaLPNcBa+OePsPxSa33ijbdxvML2kF48ARB0OVUSHG5wMwmS7oiDYgrprgA66VFQ3EzzwepnOJyqpm1SHHB07JGR4pL0AHKHiYf4AOAzfB2B975QtLlcu+o1YFxwEmSzofmjnv2Wem4I4FV0nFeLH40BsxBo5H0D2v2KTsqteFrigFHdMqDnkSFPsSRwMdmdkG1KJ4k7hyIm4wfmUWqzYdXpCobL28k6dG0307A/JIuaeR15aSJgM/iA+pbzGxsSp9pUrPB7rfxaImgFyLpeTM7KX29w0qePKVtjzaz/+GpVAfgqYXv4EbHZwK/lPRWO85hmpk9m76OMLOZeDn2LwO/kntDFX6Y75v77vwL2BefZH0XuKFIzZV0UzrW2Px3IrKza4kUrSAIupRIcQl6MxU8TFbGjRNzkQc8THkcLtT8UNLZafuKnZw0a/fHdJz38DKlr5S3C4JG0hnpWkHQSLI+hPCJ6iE0R9qcApwn6ZPSPoW4cyFwuKSLrKXPVKUS6hOBTXER6V5glKTbG3ltZcxsFC7w3IVPUE1Nyw0vSvGWpB905TkFtUlR6DOTQFeXqNGWdC0zOxRvxwCnA2Mlvd8J5z0UN+xeGa94NUnSPaX7pOiXDwa+ifd9PgLWlvROdqy8+EkIO91ApGgFQdBlRIpL0NspxJ2sA/MMMIbZ2/IN+CDkMElnm1lTLXEndeTeTIuuC3En6A46I10rCBpFqQ/xNTy9ZCbNpZhPAo42s7mzfeagJO6kVbLmsuhj8MEytIzkGYenomzd1eJOOq9b8cmr24DtzeyrZvYd4EfA/xXiTtyb3Y+ZDUkfhxbv+daEjSSUFOlap+A2AneY2VbVnrmp/R6Vvo7vDHEnHXcy7m/5DD7RtHpaPiPbZmbqx0zFTc0vwi0Sdk/XM0vYqff/QdAYIkUrCIKuJFJcgj5B3qYkPWNmY9LXoi3PxN+xReerKIc+W1tMnaZl8Lz2acAVDTz1IKhJW9K1rI6qWkHQiewN/AS4QNIHZvYJcB6wBv58NVzkwczOxp+3+1MSd6ylKXilEupz0Bxl3C0VN4tzlHSumS0BLA/MA/wdeFjSk+VrCbqHlOb3hJlNBqaZ2fPAB7hQ/gwwFfgYrxg7DXg9pTNNLY4h6V9mdgze7m43s60l3V0lXeuCJP482ZnXIWmymT2N32OLVNlG6ZymmNnpeCTPSsW6zjyfoP1EilYQBF1KpLgEfZUabflgST9O28zWWUuh0SfiPhAbSfpLF51yEFSllK41VtLVafkQYHlJT3fn+QX9EzMbJmli9v0reCXONfAJoyLq4Tw8Euckqog72THystB5ulaniTt536UNqTs1twtxp/sxs3nx6lbFxE61aCrhos8AXOj5BBdz/gZ8CLyEe9usifvbrAuMlPTnrv53rqetFn15M/s+PrG6awg8PYcQeIIg6BZKL5BqA+PpwHclXWk1SpZmecE/A/YELpF0aOOvIghaUqMt5/5SeU77QOCreNWJ7STd28WnHARVSSLPubjIcwZwIz74uARYI0SeoKsoDzRLUTjVRB6A4ySdm7ZrNeLMzLbAJ4s2lvTnzjrfoO9hLSu5VULAJNwIufg+CO/bDsyWWenzp+mYM4FNJf2pLSJPe8TEtlIcNwmiC0r6dmf/RtB+ImczCIJuoZziwuw+JtVSXCodK1Jcgh5BlbYMLf2lcnFnJ+BXuIFmiDtBj0LuyXMM8B9c4LkTuAwXI0PcCbqM8iA1H+xKGoe300dp9uMp/s4ws7nTgLSquGPOvsDvcM+djog7TSUxalszO9fM7jKz5QuvkqD3kvygXsX7nO+kzyfjfkl/xaN6puEROwPTf4PS7rlFyvT03zQ8lUu4IDQRH6ffZ2Ybqk4ftKLtZf5SHRJ3zGywmW1hZotny3LR6EO8vxN+lz2IiOAJgqDHECkuQV+htUie1Pn6Om5UuL2ku7rhNIOgKqVZ4K3waj6DgM0k3d+tJxf0W8pROKV2mkfy5IPo04BzVKquVeHY9wAXyc2N23JORRRxEdUwF/7c/wGwHF6NaxLwnKQ123LsoOeRiTvn49VdTwIWwCu8/irbZgguNI4APgMshlepasLNiT+b1i+AR+wUvk+VqDuizMwGAb8EruroxJGZHYhXbDsZv4cmZev2wPsxB6odJdqDxhEmy0EQ9BhU2awW4NLUb7pcWel0mBUFsQ1wNLBFiDtBT6BGW74szXK9SIg7QQ8mGzTPBSyBizvdYjgb9C+SAL4wnnI9Dx7J+6Kkm+WVNWd55pSigceZ2dG4yLMmPlg2fAJIZnaepI9r/PRX8wFsHedZGCEXk05LmtkmwAH4AP41vDLS1rgQMCFSt/oE5YIhn+J+ZWOSR9RVkt7L2ukdlQ5iZnPibfTzwHBc5FkWmCt9XwlvvyvRMiK4ItkE6GrAF8mKjXSAj/F2/EJJ3Pkanl7+HUlvhR9UzyIieIIg6HHU42OStitSXK4lBspBD6RGWwbYRtIfu/6sgqB+zOyr+CB1W0l3d/f5BH0bMxuGPzO3Bb6QrfoQuENSUZK5qp+Oma3H7CIP6bizRJ4kts8DTFRW/KGV8ytXkvsM8CXgcDxKYyngYuAZpdLqZrYqcD+wgaRnQ+Tp/djsBUNyU/rzJV2TlucTkk1Ak6TpdfpBFUbG80r6sA3ndgEuvqwi6dO2X91sx9sbmAxsgt9L/wEWB34g6c0Qd3oeIfAEQdAjiRSXoK+Q2vIPgc3xmTnooIFnEHQVZjYCWEDSPd19LkHfJpnWPohHLDThniSDcX+SJlysuUfSlnUcq5bIcybue/J1vIz6LpLebuV4A1L0UJGGNQoXoE4Ank/nejTwjqTxpX0PSb+zJfDfEHf6DiWz72oiT1sMkvPjtUkITILlEsAtwBmSbuyI+FJBzFwYGCTpP5n4FOJODyQEniAIeiy1PHnw0NNbCHEn6KGUvCFWB67H/RgizSUIgiDDzIYDLwP/wCPGngZWAA7FfUum4Qa0M3ET5FafoTVEnrPwtJOxwM5FpE0rxxqKe6p8Dy8LvTluOv4n4FpJ71bZb2HgCXzAf35rvxP0bjpD5Ong72+I+6WNlPREA38nhJ0eTAg8QRD0aCLFJejtJMPDzXFBcnNJD3TzKQVBEPQYkrgzHrga9zUZn60bjHuYjMRFnkHAnoWZbR3HzkWecgn1LeuJTEsm4ycAK+Ii1EPARcCrhRdQ2i6PviiMl5fBB9y7SXo60rP6Pt0h8qT0r8HAzcDbkvZqxO8EvYMokx4EQY9GzWWn7wLyChgbh7gT9BK2xEvvbhfiThAEQTOZuHM5cFIh7pgzUNJUYCvgGZrLTM+XbWPpc8UxjaSHgWPxyKCp2aqN6xR3huAmuHfiKeJrAEdKehmPKJpVHlotS7cXn88BPpT0dFoe4k4fR9I/8Db3NnCUme2XltdV6rydvzlT0mRgUTzNseo9EfR94h8+CIIeS9ZpegY4HXgzrdok/EuCXsRbwKgQJIMgCJrJxJ0rgdMkfVSskzM9eX1MBy7BBZoZpOpASSyZP+0yIB3TsuMXfYhxwPHA62lV3X0ISVPkFTzPkHRDad2M7DwqXV9Rge6C9L1aCeyK5NcS9C66Q+QxsxPw0uvXFr/ViN8Jej4h8ARB0GPJ/EsGAYsAywCbhn9J0JuQ9GgY1AZBEDSTDJXH4+lTJ1erEpRVKnoST0F5HnjEzE40sxuBR81sHDDWzEYkA+RC2Mn7EHPjFa7a3IcoCy1tGDjvgBtGP5X2q1k1Kfu9edP2Ee3Ti2mLyNNW8a+0b9E+vwD8Ih2/3ccLej8h8ARB0BuIFJcgCIIg6AMkcec1YE48femjtLzWoHTh9PdJ4KfAKcDOePrUunjxhUfMbN0KwshWuHFzu/oQbRVasuvYBi/L/n/1Rm2Y2TeB68zss208zaAHUkHk+WZaPtPMhpjZKul7XeJfld+Qma0B7ImnEnboeEHvJwSeIAh6A5HiEgRBEAS9HDObn2ZxB+ACM/sW+KC0QmRD8X299Hco7sXzIPDvbNOJeJrWqWa2aOlnlwFGd1UfIl3Hgvjk1L/TslajflKEx1XAryS92dr2Qe8gE3newUWePVJU2d54BNoq7T12Jiaug3tV/qme1D4zG9je3wx6PlFFKwiCIAiCIAiChpI8d17BfXQGAHNlq78j6fK0XVGBakASS9YHHsArWZ2bpV6tgA+Sj0vHmIkPoreS9FSXXFSJ7NxPSOexQZ37fRMXd3aXdH1U2+p7pOpapwMrAC8AG+PVYO/uhGP/G7hL0mFV1luK9BmQR/eY2deAxYCJkq7o6HkEPYOI4AmCIAiCIAiCoGGY2QLABNxweBRwLx51U3CZmR0Es9JXBidxZ1XgHuB4SeekQWpT2u7fkr6HV6oCH9csDGyS/W6XGhVnkTobAs+a2YDW0rMycWePQtxp9HkGXUfmCfUPYCzeRkcCW3RE3MkqyG0KTEnHLhuNF/eKsmXLmtmZZvYYcDEwmuQTFfQNIjwrCIIgCIIgCIJGsjReLetCSe+b2Qe4ILMZMCxtc5mZkapWTTWz1YFxePn082FWJMLM7LPScUcCRarLi8WPdkcUjJltA6wF7N2aF0pJ3LmubBAd9H6yiLOiqtogvJJbhwqGZG1kOzwq7r0K28xMbWoO4LvAV9L29+NRcScD0yVNiqixvkMIPEEQBEEQBEEQNAxJj5rZU0VVLEn/NLMxaXVZ5HkPF3bGAT+UdC60EHSKYxafxwMf4FW23gP+1ejrqUSW/rIaHhHxiZkNTGXeK20f4k7/YiRwKZ661ynVYM1sSWBz4BRJH0MLQWkpYEE8wm1uYFk8yufHuR9VHuVTvsdC9OmdhMATBEEQBEEQBEFDKcSdYtAo6ZkqIs8NwHTgu5KuTANQVRpoZp43hSnxdZJeafClVCSllM0PHAmcJumTatuGuNMveR333Lmnowcq2j2wHDAceDpbtxGeprgf8ElaNxa4W9KU0nFmRcTBLJFnCLChpHuiPfZOQuAJgiAIgiAIgqBLKEXhVBJ5ZuJjlPeLzQBLf8vHmmlmy+BVhKYB3W0UuxIeTXQrVI6ACHGnfyLpsU481sxUQetE4A7gfTM7EPh22uRT4BjgaUnPFfuVI8oqtM1FgT8Dt+HeV0EvJASeIAiCIAiCIAi6hRqRPDea2YKSfgwoi1qYhZkNBfbFIxk2ktQt6VkZ8wAPSXoJKg6gQ9wJOot58LSrJYE3cV+dx3DR59MiZQuaI36qpQumbT6HizovV6vGFfQOokx6EARBEARBEATdipmtDJxKS5EHWpZQn1Xm2cwGAl8FrgO2k3RvF5/ybJjZ/7d398GblXUdx9+ffRAh3GWVR3WZZFVUAlxdXJMshaHNILWH3WFklCkdWqKppoeZ0lUREqipZpxp0NEUBKcwCxUxxkQ0NqAaoiG2EkUYMiRkWxfjIV3g2x/nurezdz+WRXd/931+9/v1z57fOdd939eZObMz5zPX9f0ur6oH2vF4S2rDHe01re36RXShznX9ujrfw3c9my4guruq1rVz/y9Q1TAY8EiSJEmauD0Jedq4JcDPApcCb6yqz83nPJ/MHMVqDXe0VyXZb7ymTjv/lAojG+4sPIsmPQFJkiRJqqrbgHcC1wIP9y5dnGQjdCtjgPVMabgDuwY3SX6JrjaQ4Y72mlG4M3qeeucNd2acK3gkSZIkTY3drOQ5B7gDuIopDXf6kjwX2EzX7v2jhjuaFoY7C5cBjyRJkqSpspuQB7p2099zzZH50tqmH1xVXzXc0bQw3FnYDHgkSZIkTZ0W8pwLnAIc2E6/pqqun9ikpAEz3Fn4rMEjSZIkaWr0VrvcBryXrg00wGvnI9xJsnR8LuPH0tAY7swGAx5JkiRJU2O0jakFLYcDq4CTq+pv9vVvJ3kBsLZ16qKqKskz+/OShsZwZ3YY8EiSJEmaRuuAK4HTquqL8/SbdwJHAr8GkORg4D1Jjpun35f2qhbuXIvhzkywBo8kSZKkqZNkDbCiqj4/gd++FfgacDvwmaq6cb7nIH2/kiwH/gXYUlU/0c4Z7ixgruCRJEmSNHWq6ub5DndGW7OAS+mKO584CneSLG7/7pfk0Pmcl4YlybS8Z59AF1Aa7syIaXnwJEmSJGmiqurRFuQ8AJwKPCfJX7Rrj7VhBwGvS3LIhKapKVdVjyd5WpJ3J3n+BKeyuarOBsOdWeEWLUmSJEkzr/8CnGRlVX09yQF0W1xurqr1vbGr6Oou3zmh6WrKJXkJsAV4f1Wd084t7gWF0l7nCh5JkiRJM6+/uqGqvt7+fRg4BliT5BO94QGOnt8ZamC+AdwEnJ3kLfB/q8CSHDbJiWnhcgWPJEmSpJmW5FnAemAFsD/wd8AXquo77foBwC3APwPvBJ4HPFBVN01mxhqCJBuAK4B7gZ+qqluSnA0cBWwaPV978D3HA/cAB1XVHftswho8V/BIkiRJmlmt09AJdKHOnwLLgTOBTyc5EHau5Hkp8AiwATjKcEdPJEkAqurPgd8DjgBemeRXgd8APrS7cGf0+Xb848Aq4DnARUlW78u5a9hcwSNJkiRpJiVZBrwVuLqqvtrOLQV+BPhDum02P1NV323XUr5AaQ+MajolORa4hG6r353Aj1XV1j2px5PkbcB/VdUnk7wYOAP46OhZlca5gkeSJEnSrFoGfKsX7iyqqh3A9cD5dNu1jutdM9zRU1JVt9EVW94P+HZVbW2X9qSj1bKq+mT7nn8D/gD42j6ZqBYEAx5JkiRJC1aSJb3jjF3+AeD4dm1pW3GRtrLiamAH8BrYtQiz9GR6Hdl+AXgFXTe2tUnOaat3dhsWtrpQZ7SObaPv3D56RpOsTPJD+/IeNDwGPJIkSZIGZTyomSO4GZ1/IfCKJIuh62ueZEVvyFLgTUmOqqodoxfvFvbsAC4EvryPbkMLXNtidS5wMvCTwFbgdLqVYSRZ1Bu7uHcc4L+B+2jd2vpBZQuHvgk8d1/fg4bFgEeSJEnSYPS3SvWDm/b3eNBzF12B2l9u1w8BzhutfKiqLcB1wHVJVlbVY71wB7puWbutkyLNpT2LW4GTqupe4H+ATwMnAu+AbpVPkkUt6Fmf5OXtfLW6T7cDH0nyjKp6dOz5Pgx4dB5vSQNgkWVJkiRJgzAqXNuOP0jXWejpwMXANa3b1fhnFgO30rU4vwv4q6q6oVcE9zjgfcCLgXXA3VW1PclPAw9W1efn5ea0YI0KKidZA3wJOICuePenemNWA6dU1e8nWVJVj7bzm+me8w3Av1bVw0mOAZ4PbK6qbfN9P5peruCRJEmSNAi9cOc8uhURN9EFM+8D3tDfxtLGLW31dC4D3gCsraob2uXRu9BXgLcDlwMfAy5J8uvAoaNw54m2gEl7YtQtq6puBja20+uTHNTq6SwBHgZGK8v6K3NOB24GrgGuT/KbwFnA3xruaJwreCRJkiQNRpILgXuq6o/b36fShTNfBk6tqm/1V/q0MWfRtai+BPhSVb25nV8EHAKsq6rLkhwGBNhma3TtTaPnKMnBwEeA1wGvr6premPeC5zb2yLY//zL6Vb+3A08UlX3z9PUNSCu4JEkSZI0JP8xCmbKXOcAAAR6SURBVHeavwc+B7wS2AQ7a5sckuTn2pjPVtW1wLHASUkuH42rqvuAf0jyvKq6r6r+03BHe9voOWpt0j8BLAY+nOTo3rDHgPFVaKM6U/9YVZur6t8Nd/REDHgkSZIkTaV+l6H294uANe14Kex8Yb4I2AZsTPLD7fz9wIoky6rqnrZdaztwDL2Qp3kceOH47xvuaG8abfWrqsuBDwKHA29PsjzJq4FDGSvqPdreJe0JAx5JkiRJU2esoPKhSZ4NPADcMTYuVXUr8Dt0BZfXtvNPA44EfhSgtUFf0gt5XpXkY0lWAT8IfHt+7kyzqm3RWtSONwJ/DbyZbgXaz9Ntz/ruBKeogbMGjyRJkqSpMhbufAA4AVgNXA/cXlW/OMdnVtO9MO8HrKmqryQ5HjiurZgYjVvSWk4vA95PV2T5/qq6eJ/fmMSuW/+SvAV4CLixqu4drx8lPRUGPJIkSZKmUpJ3AdvpWpyfDpxJ1+p8Ld1LcfW3USW5APht4N1VdX6rb3LyeHjTa1vty7QmYq76TqPnclJz0vAtefIhkiRJkjS/Wuerbb1uWVuBlXTdh06rqj/rjR0FNb8LnEJXi+cGYB1wy/h399pW9zttWVBZ82auZ81wR98va/BIkiRJmkYPAjtDnKraAoxW4nw4yct610ZBzXeAC4FvAOcB26vq43vyY4Y7kobOgEeSJEnSRI1aQbfjJNkfeCmwS+hSVZ8FLqArpnxmkqePOhO1649V1ZV0W7jOqKoL23f63iNpwbMGjyRJkqSJGSs4+w7gCLpiyq8CNlXVg+3aoqp6PMmBwFXAi4ATq+qu3dUusc6OpFlhki1JkiRpYnrhznuAo+iCmyuA04DlvXGjkOYh4ErgcOBPkixtBZPDHAx3JM0KAx5JkiRJE5XkCGBHVb0V+C3gRrqVPM9s13e+t7RA6FLgNuC1wFkWSJYkAx5JkiRJk3c/3W6tJcAWui1Y+9Ot0Fkx1u1qcdu29Sbgm8AGfK+RJP8jlCRJkjRxAZbStT/fAfwRcA1wAvC2fhHm3nasu4DLgFcDJ8//lCVpulhkWZIkSdLEJTka2AS8qxVOPhK4CbgTOKmqdowXTE6yCjimqq6azKwlaXoY8EiSJEmaCklWAtuq6qEkzwDOB34FuKCqNrUxc9bb2V0nLUmaBQY8kiRJkqbCeHiT5HjgOuAg4I1V9ZmJTU6Sppw1eCRJkiRNhfGVOVV1K7CRrkbPhiTLnqgduiTNOgMeSZIkSdPsauAvgdOBl9kOXZLmZsAjSZIkaWpV1SPAFcA9wLETno4kTS1r8EiSJEmaSv2aPElWV9U/TXpOkjStDHgkSZIkTa05Ci/v0ipdktQx4JEkSZIkSRo4a/BIkiRJkiQNnAGPJEmSJEnSwBnwSJIkSZIkDZwBjyRJkiRJ0sAZ8EiSJEmSJA2cAY8kSZIkSdLAGfBIkiRJkiQNnAGPJEmSJEnSwBnwSJIkSZIkDZwBjyRJkiRJ0sAZ8EiSJEmSJA2cAY8kSZIkSdLAGfBIkiRJkiQN3P8C+nJV2xygd5wAAAAASUVORK5CYII=\n",
"text/plain": [
"