def plot_the_pred_distribution_for_four_systems(model, bonds_new_df, bar_ene):
# features for four systems
data_wt_d1 = bonds_new_df[0:200]
data_wt_d2 = bonds_new_df[200:400]
data_y72_d1 = bonds_new_df[400:600]
data_y72_d2 = bonds_new_df[600:800]
# energy for four systems
y_wt_d1 = bar_ene[0:200]
y_wt_d2 = bar_ene[200:400]
y_y72_d1 = bar_ene[400:600]
y_y72_d2 = bar_ene[600:800]
y_wt_d1_pred = model.predict(data_wt_d1).flatten()
y_wt_d2_pred = model.predict(data_wt_d2).flatten()
y_y72_d1_pred = model.predict(data_y72_d1).flatten()
y_y72_d2_pred = model.predict(data_y72_d2).flatten()
wt_d1_error = y_wt_d1_pred - y_wt_d1
wt_d2_error = y_wt_d2_pred - y_wt_d2
y72_d1_error = y_y72_d1_pred - y_y72_d1
y72_d2_error = y_y72_d2_pred - y_y72_d2
# print("MAX error in the training set:", max(train_error))
# print("MIN error in the training set:", min(train_error))
# print("MAE for the training set:", np.mean(np.abs(train_error)))
rstr_r2 = r'$R^2$'
rstr_mae = r'$MAE$'
rstr_rmse = r'$RMSE$'
rstr_nsamp = r'$n_{samples}$'
# WT_D1
r2_wt_d1 = round(r2_score(y_wt_d1, y_wt_d1_pred),4)
mae_wt_d1 = np.mean(np.abs(y_wt_d1 - y_wt_d1_pred))
rmse_wt_d1 = (np.sum((y_wt_d1 - y_wt_d1_pred)**2)/y_wt_d1.shape[0]) ** 0.5
# n_train = y_train.shape[0]
# WT_D2
r2_wt_d2 = round(r2_score(y_wt_d2, y_wt_d2_pred),4)
mae_wt_d2 = np.mean(np.abs(y_wt_d2 - y_wt_d2_pred))
rmse_wt_d2 = (np.sum((y_wt_d2 - y_wt_d2_pred)**2)/y_wt_d2.shape[0]) ** 0.5
# Y72_D1
r2_y72_d1 = round(r2_score(y_y72_d1, y_y72_d1_pred),4)
mae_y72_d1 = np.mean(np.abs(y_y72_d1 - y_y72_d1_pred))
rmse_y72_d1 = (np.sum((y_y72_d1 - y_y72_d1_pred)**2)/y_y72_d1.shape[0]) ** 0.5
# Y72_D2
r2_y72_d2 = round(r2_score(y_y72_d2, y_y72_d2_pred),4)
mae_y72_d2 = np.mean(np.abs(y_y72_d2 - y_y72_d2_pred))
rmse_y72_d2 = (np.sum((y_y72_d2 - y_y72_d2_pred)**2)/y_y72_d2.shape[0]) ** 0.5
# plot a figure for the predition and true value
sns.set_style(style='white')
sns.set_style("ticks")
fig, axs = plt.subplots(2,2,figsize = (7,7), dpi =300, facecolor = 'w', edgecolor = 'k')
# KPC-WT/IPM-D1
plt.subplots_adjust(left=.3/7, right=6.7/7, top=6.9/7, bottom=.1/7, wspace = 0.3, hspace = 0.3)
axs[0,0].tick_params(which='major', direction='in', length=6, width=1)
axs[0,0].scatter(y_wt_d1, y_wt_d1_pred , color=sns.color_palette('colorblind')[0], zorder=2, s = 3)
axs[0,0].axis([-5, 85, -5, 85])
xpoints = ypoints = axs[0,0].set_xlim()
axs[0,0].plot(xpoints, ypoints, linestyle='-', color='r', lw=1, scalex=False, scaley=False)
axs[0,0].set_title(r'KPC-WT/IPM-$\Delta$1')
axs[0,0].set_xlabel(r"$\Delta E_{QM/MM} (kcal$ $ mol^{-1})$")
axs[0,0].set_ylabel(r"$\Delta E_{ML} (kcal$ $ mol^{-1})$")
axs[0,0].xaxis.set_major_locator(plt.MultipleLocator(base=20.0))
axs[0,0].yaxis.set_major_locator(plt.MultipleLocator(base=20.0))
axs[0,0].text(0.94, 0.06,
f'{rstr_r2: >10} = {r2_wt_d1: <8.4f}\n{rstr_mae: >10} = {mae_wt_d1: <8.4f}\n{rstr_rmse: >10} = {rmse_wt_d1: <8.4f}\n {rstr_nsamp: >10} = {200: <8}',
ha='right', va='bottom', transform=axs[0,0].transAxes, fontsize=10)
#. KPC-WT/IPM-D2
axs[0,1].tick_params(which='major', direction='in', length=6, width=1)
axs[0,1].scatter(y_wt_d2, y_wt_d2_pred , color=sns.color_palette('colorblind')[0], zorder=2, s = 3)
axs[0,1].axis([-5, 85, -5, 85])
xpoints = ypoints = axs[0,1].set_xlim()
axs[0,1].plot(xpoints, ypoints, linestyle='-', color='r', lw=1, scalex=False, scaley=False)
axs[0,1].set_title(r'KPC-WT/IPM-$\Delta$2')
axs[0,1].set_xlabel(r"$\Delta E_{QM/MM} (kcal$ $ mol^{-1})$")
axs[0,1].set_ylabel(r"$\Delta E_{ML} (kcal$ $ mol^{-1})$")
axs[0,1].xaxis.set_major_locator(plt.MultipleLocator(base=20.0))
axs[0,1].yaxis.set_major_locator(plt.MultipleLocator(base=20.0))
axs[0,1].text(0.94, 0.06,
f'{rstr_r2: >10} = {r2_wt_d2: <8.4f}\n{rstr_mae: >10} = {mae_wt_d2: <8.4f}\n{rstr_rmse: >10} = {rmse_wt_d2: <8.4f}\n {rstr_nsamp: >10} = {200: <8}',
ha='right', va='bottom', transform=axs[0,1].transAxes, fontsize=10)
# KPC-F72Y/IPM-D1
axs[1,0].tick_params(which='major', direction='in', length=6, width=1)
axs[1,0].scatter(y_y72_d1, y_y72_d1_pred , color=sns.color_palette('colorblind')[0], zorder=2, s = 3)
axs[1,0].axis([-5, 85, -5, 85])
xpoints = ypoints = axs[1,0].set_xlim()
axs[1,0].plot(xpoints, ypoints, linestyle='-', color='r', lw=1, scalex=False, scaley=False)
axs[1,0].set_title(r'KPC-F72Y/IPM-$\Delta$1')
axs[1,0].set_xlabel(r"$\Delta E_{QM/MM} (kcal$ $ mol^{-1})$")
axs[1,0].set_ylabel(r"$\Delta E_{ML} (kcal$ $ mol^{-1})$")
axs[1,0].xaxis.set_major_locator(plt.MultipleLocator(base=20.0))
axs[1,0].yaxis.set_major_locator(plt.MultipleLocator(base=20.0))
axs[1,0].text(0.94, 0.06,
f'{rstr_r2: >10} = {r2_y72_d1: <8.4f}\n{rstr_mae: >10} = {mae_y72_d1: <8.4f}\n{rstr_rmse: >10} = {rmse_y72_d1: <8.4f}\n {rstr_nsamp: >10} = {200: <8}',
ha='right', va='bottom', transform=axs[1,0].transAxes, fontsize=10)
# KPC-F72Y/IPM-D2
axs[1,1].tick_params(which='major', direction='in', length=6, width=1)
axs[1,1].scatter(y_y72_d2, y_y72_d2_pred , color=sns.color_palette('colorblind')[0], zorder=2, s = 3)
axs[1,1].axis([-5, 85, -5, 85])
xpoints = ypoints = axs[1,1].set_xlim()
axs[1,1].plot(xpoints, ypoints, linestyle='-', color='r', lw=1, scalex=False, scaley=False)
axs[1,1].set_title(r'KPC-F72Y/IPM-$\Delta$2')
axs[1,1].set_xlabel(r"$\Delta E_{QM/MM} (kcal$ $ mol^{-1})$")
axs[1,1].set_ylabel(r"$\Delta E_{ML} (kcal$ $ mol^{-1})$")
axs[1,1].xaxis.set_major_locator(plt.MultipleLocator(base=20.0))
axs[1,1].yaxis.set_major_locator(plt.MultipleLocator(base=20.0))
axs[1,1].text(0.94, 0.06,
f'{rstr_r2: >10} = {r2_y72_d2: <8.4f}\n{rstr_mae: >10} = {mae_y72_d2: <8.4f}\n{rstr_rmse: >10} = {rmse_y72_d2: <8.4f}\n {rstr_nsamp: >10} = {200: <8}' ,
ha='right', va='bottom', transform=axs[1,1].transAxes, fontsize=10)
# fig.suptitle(f"SVM Model", fontsize = 12)
plt.tight_layout()
fig.savefig(f'./SI/2.SVM_four_systems_results.png')
