import warnings warnings.filterwarnings("ignore") import math import sys import csv import numpy as np from netCDF4 import Dataset from decimal import * from tqdm import tqdm import matplotlib import matplotlib.pyplot as plt import os import accpostproc as accp from PIL import Image import lib import regionfinder import argparse lib.plt_rcParams_sci_adv() parser = argparse.ArgumentParser(description='') parser.add_argument('--regions', type=str,nargs = '+',default =["CHN","ASEAN","E-ASI","EUR","NAFTA"]) parser.add_argument('--runmean', type=int, default = 14) parser.add_argument('--starttime', type=int, default = 0) parser.add_argument('--numbering', type=str, default = None) parser.add_argument('--numbering_position', type=float,nargs='+', default = [0.03,0.95]) parser.add_argument('--ylower', type=float, default = -3.95) parser.add_argument('--yupper', type=float, default = 3.95) parser.add_argument('--yticks', type=float, nargs='+', default = [-2.5,0.0,2.5]) parser.add_argument('--endtime', type=int, default = None) parser.add_argument('--extra', type=str, default = '') parser.add_argument('--scenario', type=str, default = "WPTS") parser.add_argument('--variable', type=str, default = "export") parser.add_argument('--relation', type=str, default = "perc",choices = ["abs","perc"]) args = parser.parse_args() scale = pow(10,6.) seasons = [2000 + i for i in range(21)] networks = [2000 + i for i in range(16)] d = Dataset(f"blocs_{args.scenario}-0p25_kl-22_en-2015_2018-05-01-2019-01-01_ar-120_ul.nc") time = d.variables['time'][args.starttime:args.endtime] blocs = d.variables['region'][:] number = math.ceil(len(args.regions) / 2.) ix = [0,1] ids = [0,2] arr = np.array([[[1,2,3],[4,5,6],[7,8,9]],[[10,11,12],[13,14,15],[16,17,18]],[[19,20,21],[22,23,24],[25,26,27]]]) correct_ones = lib.get_rows_in_file("correct_finished.csv") fig = plt.figure(figsize = (3.5,0.5+1.*number), dpi = 600) print(f"Make ---- {args.variable} Plot") quantities = np.full(shape=(len(networks),time.shape[0],blocs.shape[0],blocs.shape[0],len(seasons)),fill_value = np.nan) values = np.full(shape=(len(networks),time.shape[0],blocs.shape[0],blocs.shape[0],len(seasons)),fill_value = np.nan) zeros = np.full(shape=(len(networks),blocs.shape[0],blocs.shape[0],len(seasons)),fill_value = np.nan) for num_season,season in enumerate(seasons): for num_network,network in enumerate(networks): name = f"{args.scenario}-0p25_kl-22_en-{network}_{season}-05-01-{season+1}-01-01_ar-120_ul" if name in correct_ones: d = Dataset(f"blocs_{name}.nc") time_shape = d.variables['time'][:].shape[0] quantities[num_network,:,:,:,num_season] = (d.groups[f"regions"].variables[f"trade_quantity"][args.starttime:args.endtime,:,:] - d.groups[f"regions"].variables[f"trade_quantity"][0,:,:])/ scale values[num_network,:,:,:,num_season] = (d.groups[f"regions"].variables[f"trade_value"][args.starttime:args.endtime,:,:] - d.groups[f"regions"].variables[f"trade_value"][0,:,:] )/ scale zeros[num_network,:,:,num_season] = d.groups[f"regions"].variables[f"trade_quantity"][0,:,:] / scale d.close() q_positions = [n-0.2 for n in networks] v_positions = [n+0.2 for n in networks] for num, region in enumerate(tqdm(args.regions,desc="Main",leave = False)): ax = plt.subplot2grid((number,2), lib.number_to_row_of_two(num),colspan=1,rowspan = 1) if num % 2 != 0: ax.set_yticklabels([],fontsize = 0) ax.axhline(0,color = lib.colordict['gray']) ix = lib.subgroup_index(blocs,region) if "trade" not in args.variable and "port" not in args.variable : q = np.nansum(quantities[:,ix,:],axis=(1)) v = np.nansum(values[:,ix,:],axis=(1)) z = np.nansum(zeros[:,ix,:],axis=(1)) elif args.variable == "domestic_trade": q = np.nansum(np.nansum(quantities[:,ix,:,:],axis=(1))[:,ix,:],axis=(1)) v = np.nansum(np.nansum(values[:,ix,:,:],axis=(1)),axis=(1)) z = np.nansum(np.nansum(zeros[ix,:,:],axis=(0))[ix,:],axis=(0)) elif args.variable == "foreign_trade": ix_rest = lib.get_rest(lib.subgroup_index(blocs,blocs),lib.subgroup_index(blocs,region)) q = np.nansum(np.nansum(quantities[:,:,ix,:,:],axis=(2))[:,:,ix_rest,:],axis=(2)) + np.nansum(np.nansum(quantities[:,:,ix_rest,:,:],axis=(2))[:,:,ix,:],axis=(2)) v = np.nansum(np.nansum(values[:,:,ix,:,:],axis=(2))[:,:,ix_rest,:],axis=(2)) + np.nansum(np.nansum(values[:,:,ix_rest,:,:],axis=(2))[:,:,ix,:],axis=(2)) z = np.nansum(np.nansum(zeros[:,ix,:,:],axis=(1))[:,ix_rest,:],axis=(1)) + np.nansum(np.nansum(zeros[:,ix_rest,:,:],axis=(1))[:,ix,:],axis=(1)) elif args.variable == "export": ix_rest = lib.get_rest(lib.subgroup_index(blocs,blocs),lib.subgroup_index(blocs,region)) q = np.nansum(np.nansum(quantities[:,:,ix,:,:],axis=(2))[:,:,ix_rest,:],axis=(2)) v = np.nansum(np.nansum(values[:,:,ix,:,:],axis=(2))[:,:,ix_rest,:],axis=(2)) z = np.nansum(np.nansum(zeros[:,ix,:,:],axis=(1))[:,ix_rest,:],axis=(1)) elif args.variable == "import": ix_rest = lib.get_rest(lib.subgroup_index(blocs,blocs),lib.subgroup_index(blocs,region)) q = np.nansum(np.nansum(quantities[:,:,ix_rest,:,:],axis=(2))[:,:,ix,:],axis=(2)) v = np.nansum(np.nansum(values[:,:,ix_rest,:,:],axis=(2))[:,:,ix,:],axis=(2)) z = np.nansum(np.nansum(zeros[:,ix_rest,:,:],axis=(1))[:,ix,:],axis=(1)) data_q = [] data_v = [] for en,network in enumerate(networks): notnan = np.where(z[en,:]/z[en,:] == 1.)[0] z_c = z[en,:] z_c = z_c[notnan] q_c = q[en,:,:] q_c = q_c[:,notnan] v_c = v[en,:,:] v_c = v_c[:,notnan] q_en = 100 * np.nansum(q_c[:,:],axis=(0)) / (time.shape[0]*z_c[:]) v_en = 100 * np.nansum(v_c[:,:],axis=(0)) / (time.shape[0]*z_c[:]) data_q.append(q_en) data_v.append(v_en) ax.boxplot(data_q,positions = q_positions,whis = (5,95),showfliers=False,widths = 0.25,medianprops = dict(color=lib.colordict['blue_alt']),boxprops = dict(linewidth = 0.4),whiskerprops = dict(linewidth = 0.4), capprops = dict(linewidth = 0.4)) ax.boxplot(data_v,positions = v_positions,whis = (5,95),showfliers=False,widths = 0.25,medianprops = dict(color=lib.colordict['orange_alt']),boxprops = dict(linewidth = 0.4),whiskerprops = dict(linewidth = 0.4), capprops = dict(linewidth = 0.4)) ax.scatter([],[],color = lib.colordict['black'],label = lib.bloc_dic[region], s = 0) ax.legend(loc = 'best') ax.set_xlim(1999,2016) ax.set_xticks([2000,2002,2004,2006,2008,2010,2012,2014]) ax.set_yticks(args.yticks) if math.ceil(num / 2.) != number -1 : ax.set_xticklabels([],fontsize = 0) else: ax.set_xlabel("Econonmic network [year]") ax.set_xticklabels(["",2002,"",2006,"",2010,"",2014]) ax.set_ylim(args.ylower,args.yupper) if num == 0: if len(args.regions) % 2 == 0: tt = [None,None] ll = ['Quantity','Value'] tt[0], = plt.plot([],[],color = lib.colordict['blue_alt'],linewidth = 1.2) tt[1], = plt.plot([],[],color = lib.colordict['orange_alt'],linewidth = 1.2) legend2 = plt.legend(tt,ll,loc = 'center',bbox_to_anchor =(1.,1.1),ncol = 2) ax.add_artist(legend2) if len(args.regions) % 2 != 0: ax = plt.subplot2grid((number,2), lib.number_to_row_of_two(num+1),colspan=1,rowspan = 1) ax.plot([],[],label = "Quantity",color = lib.colordict['blue_alt'],linewidth = 1.2) ax.plot([],[],label = "Value",color = lib.colordict['orange_alt'],linewidth = 1.2) ax.legend(loc='lower center',handlelength = 0.8) ax.set_axis_off() if args.relation == "perc": fig.text(0.02,0.53,f"{lib.label_quantity(args.variable)} change [\%]",rotation = 'vertical',verticalalignment = 'center',horizontalalignment = 'center') else: fig.text(0.02,0.53,f"{lib.label_quantity(args.variable)} change [bn USD]",rotation = 'vertical',verticalalignment = 'center',horizontalalignment = 'center') plt.subplots_adjust(hspace=0.05, wspace=0.05, left= 0.15, top = 0.94, right = 0.99, bottom = 0.1) outputname = f"network_ms_{args.scenario}_{args.variable}_{args.relation}_0p25{args.extra}.png" fig.savefig(outputname) fig.savefig(outputname.replace("png","pdf")) exit()