def grafsmvdof(path, cont=False):
inputs = open(r'%s\modeloFONAG\inputs\t_inputs.csv' % path, 'r')
t_inputs = inputs.read().split('\n')[1].split('\r')[0].split(',')
inputs.close()
X_Labels = open(
r'%s\modeloFONAG\%s\calibracion\X_Labels.csv' % (path, t_inputs[16]),
'r')
X_Labels = X_Labels.read().split('\n')[0].split(',')
X_Labels = map(str, X_Labels)
mvd_m = genfromtxt('%s\modeloFONAG\%s\calibracion\mvd_m_RSA_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
mvd_lb = genfromtxt('%s\modeloFONAG\%s\calibracion\mvd_lb_RSA_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
mvd_ub = genfromtxt('%s\modeloFONAG\%s\calibracion\mvd_ub_RSA_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
fig = plt.figure("Seneibilidad MVD OF")
pf.boxplot1(mvd_m,
X_Labels=X_Labels,
Y_Label='Sensibilidad MVD',
S_lb=mvd_lb,
S_ub=mvd_ub)
print('Plot de sensibilidad MVD con intervalos de confianza')
plt.show()
def grafdpawnof(path, cont=False):
inputs = open(r'%s\modeloFONAG\inputs\t_inputs.csv' % path, 'r')
t_inputs = inputs.read().split('\n')[1].split('\r')[0].split(',')
inputs.close()
if cont:
X = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedXC.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
else:
X = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedX.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
# Nombres de parametros
X_Labels = open(
r'%s\modeloFONAG\%s\calibracion\X_Labels.csv' % (path, t_inputs[16]),
'r')
X_Labels = X_Labels.read().split('\n')[0].split(',')
X_Labels = map(str, X_Labels)
X_Labels_dummy = X_Labels
X_Labels_dummy.append('dummy')
KS_m = genfromtxt('%s\modeloFONAG\%s\calibracion\KS_m_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
KS_lb = genfromtxt('%s\modeloFONAG\%s\calibracion\KS_lb_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
KS_ub = genfromtxt('%s\modeloFONAG\%s\calibracion\KS_ub_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
fig = plt.figure(
"Sensibilidad Dummy PAWN OF") # plot main and total separately
pf.boxplot1(KS_m,
S_lb=KS_lb,
S_ub=KS_ub,
X_Labels=X_Labels_dummy,
Y_Label='KS')
print('Plot de sensibilidad PAWN con variable dummy')
plt.show()
def grafspawnof(path, cont=False):
inputs = open(r'%s\modeloFONAG\inputs\t_inputs.csv' % path, 'r')
t_inputs = inputs.read().split('\n')[1].split('\r')[0].split(',')
inputs.close()
if cont:
X = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedXC.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
else:
X = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedX.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
# Nombres de parametros
X_Labels = open(
r'%s\modeloFONAG\%s\calibracion\X_Labels.csv' % (path, t_inputs[16]),
'r')
X_Labels = X_Labels.read().split('\n')[0].split(',')
X_Labels = map(str, X_Labels)
KS_median_m = genfromtxt(
'%s\modeloFONAG\%s\calibracion\KS_median_m_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
KS_median_lb = genfromtxt(
'%s\modeloFONAG\%s\calibracion\KS_median_lb_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
KS_median_ub = genfromtxt(
'%s\modeloFONAG\%s\calibracion\KS_median_ub_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
KS_mean_m = genfromtxt(
'%s\modeloFONAG\%s\calibracion\KS_mean_m_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
KS_mean_lb = genfromtxt(
'%s\modeloFONAG\%s\calibracion\KS_mean_lb_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
KS_mean_ub = genfromtxt(
'%s\modeloFONAG\%s\calibracion\KS_mean_ub_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
KS_max_m = genfromtxt(
'%s\modeloFONAG\%s\calibracion\KS_max_m_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
KS_max_lb = genfromtxt(
'%s\modeloFONAG\%s\calibracion\KS_max_lb_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
KS_max_ub = genfromtxt(
'%s\modeloFONAG\%s\calibracion\KS_max_ub_PAWN_OF.csv' %
(path, t_inputs[16]),
delimiter=',')
fig = plt.figure("Sensibilidad PAWN OF")
plt.subplot(131)
pf.boxplot1(KS_median_m,
S_lb=KS_median_lb,
S_ub=KS_median_ub,
X_Labels=X_Labels,
Y_Label='KS (median)')
plt.subplot(132)
pf.boxplot1(KS_mean_m,
S_lb=KS_mean_lb,
S_ub=KS_mean_ub,
X_Labels=X_Labels,
Y_Label='KS (mean)')
plt.subplot(133)
pf.boxplot1(KS_max_m,
S_lb=KS_max_lb,
S_ub=KS_max_ub,
X_Labels=X_Labels,
Y_Label='Ks (max)')
print('Plot de sensibilidad PAWN con intervalos de confianza')
plt.show()
def rsa(path,
E_cond=[True, False, True, True, False, True],
E_umbral=[0.5, 1, 0.7, 25, 0.5, 1],
cont=False):
print('ANALISIS DE SENSIBILIDAD REGIONAL PARA FUNCIONES OBJETIVO')
start = time.time()
print(' ')
# %% Paso 1a: Cargar datos del proceso de calibracion
print('Paso 1a: Cargar datos del proceso de calibracion')
inputs = open(r'%s\modeloFONAG\inputs\t_inputs.csv' % path, 'r')
t_inputs = inputs.read().split('\n')[1].split('\r')[0].split(',')
inputs.close()
print('Lectura de inputs')
# Crear carpeta de salidas
print('Crear carpeta de salidas')
os.system('md modeloFONAG\\%s\\sensibilidad' % t_inputs[16])
# Matriz de combinaciones de parametros
if cont:
XC = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedXC.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
else:
X = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedX.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
print('Lectura de combinaciones de parametros')
# Nombres de parametros
X_Labels = open(
r'%s\modeloFONAG\%s\calibracion\X_Labels.csv' % (path, t_inputs[16]),
'r')
X_Labels = X_Labels.read().split('\n')[0].split(',')
X_Labels = map(str, X_Labels)
# Matriz de funciones objetivo
if cont:
E = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedEC.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
else:
E = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedE.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
E[:, 0] = 1 - E[:, 0]
E[:, 3] = abs(E[:, 3])
E[:, 4] = 1 - E[:, 4]
print('Lectura de funciones objetivo')
print(' ')
# %% Paso 1b: Analisis de sensibilidad regional
print('Paso 1b: Analisis de sensibilidad regional')
# Definir que indicadores se usan para RSA
E_Labels = ['NSE', 'RMSE', 'RSR', 'PBIAS', 'KGE', 'OF']
if len(E_cond) == 5:
E_cond.append(False) # Funciones objetivo
elif len(E_cond) == 6:
E_cond[5] = False # Funciones objetivo
elif len(E_cond) != 6:
E_cond = [True, False, True, True, False, False] # Funciones objetivo
if len(E_umbral) == 5:
E_umbral.append(1) # Funciones objetivo
elif len(E_umbral) != 6:
E_umbral = [0.5, 1, 0.7, 25, 0.5, 1]
print('Definir umbrales de aceptacion')
# Extraer funciones seleccionadas
E_umbral_mod = [
1 - E_umbral[0], E_umbral[1], E_umbral[2],
abs(E_umbral[3]), 1 - E_umbral[4], E_umbral[5]
]
E_umbrales = []
E_cond_index = []
for cond_index in range(6):
if E_cond[cond_index]:
E_umbrales.append(E_umbral_mod[cond_index])
E_cond_index.append(cond_index)
E_mod = E[:, E_cond_index]
print('Extraer funciones seleccionadas')
# Aplicar Analisis de Sensibilidad Regional
E_max = np.max(E_mod, axis=0) <= E_umbrales
E_min = np.min(E_mod, axis=0) >= E_umbrales
if E_max.all():
print('Todas las simulaciones cumplen los umbrales')
elif E_min.all():
print('No existen simulaciones que cumplan los umbrales.')
else:
mvd, spread, irr, idxb = Rt.RSA_indices_thres(X, E_mod, E_umbrales)
print('Aplicar Analisis de Sensibilidad Regional')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\idxb_RSA.csv',
idxb,
delimiter=',',
header='',
comments='',
fmt='%s')
# Plot CDFs de los parametros
Rt.RSA_plot_thres(X,
idxb,
X_Labels=X_Labels,
str_legend=['cumplen', 'no cumplen'])
plt.savefig(r'%s\modeloFONAG\%s\sensibilidad' % (path, t_inputs[16]) +
r'\p_RSA_cdf.png',
orientation='landscape',
papertype='a4')
print('Plot CDFs de los parametros')
# Calcular intervalos de confianza de los indicadores de sensibilidad usando bootstrapping
Nboot = 1000
mvd, spread, irr, idxb = Rt.RSA_indices_thres(X,
E_mod,
E_umbrales,
Nboot=Nboot)
# mvd, spread e irr tienen tamano (Nboot, M)
print('Calcular intervalos de confianza')
# Calcular media e intervalos de confianza de MVD entre los remuestreos:
mvd_m, mvd_lb, mvd_ub = aggregate_boot(mvd) # shape (M,)
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\mvd_m_rsa.csv',
mvd_m,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\mvd_lb_rsa.csv',
mvd_lb,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\mvd_ub_rsa.csv',
mvd_ub,
delimiter=',',
header='',
comments='')
# Plot resultados:
fig = plt.figure()
pf.boxplot1(mvd_m,
X_Labels=X_Labels,
Y_Label='Sensibilidad MVD',
S_lb=mvd_lb,
S_ub=mvd_ub)
fig.savefig(r'%s\modeloFONAG\%s\sensibilidad' % (path, t_inputs[16]) +
r'\p_MVD_boot.png',
orientation='landscape',
papertype='a4')
print('Plot de sensibilidad MVD con intervalos de confianza')
elapsed_time = (time.time() - start) / 60
print('El proceso de analisis de sensibilidad RSA tomo %s minutos.' %
elapsed_time)
print('FIN PASO 1')
plt.show()
def pawn_OF(path, cont=False):
print('ANALISIS DE SENSIBILIDAD PAWN PARA OF')
start = time.time()
print(' ')
# %% Paso 4a: Cargar datos del proceso de calibracion
print('Paso 4a: Cargar datos del proceso de calibracion')
inputs = open(r'%s\modeloFONAG\inputs\t_inputs.csv' % path, 'r')
t_inputs = inputs.read().split('\n')[1].split('\r')[0].split(',')
inputs.close()
print('Lectura de inputs')
# Crear carpeta de salidas
print('Crear carpeta de salidas')
os.system('md modeloFONAG\\%s\\sensibilidad' % t_inputs[16])
# Matriz de combinaciones de parametros
if cont:
X = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedXC.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
else:
X = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedX.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
print('Lectura de combinaciones de parametros')
# Nombres de parametros
X_Labels = open(
r'%s\modeloFONAG\%s\calibracion\X_Labels.csv' % (path, t_inputs[16]),
'r')
X_Labels = X_Labels.read().split('\n')[0].split(',')
X_Labels = map(str, X_Labels)
# Matriz de funciones objetivo
if cont:
E = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedEC.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
else:
E = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedE.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
E = E[:, 5]
print('Lectura de funciones objetivo')
print(' ')
# %% Paso 4b: Analisis de sensibilidad PAWN
print('Paso 4b: Analisis de sensibilidad PAWN')
n = 10 # numero de intervalos condicionantes
# Calcular y plotear CDFs condicionales y no condicionales:
YF, FU, FC, xc = PAWN.pawn_plot_cdf(X,
E,
n,
cbar=True,
n_col=3,
labelinput=X_Labels)
plt.savefig(r'%s\modeloFONAG\%s\sensibilidad' % (path, t_inputs[16]) +
r'\p_PAWN_OF_cdf.png',
orientation='landscape',
papertype='a4')
print('Plot de CDFs condicionales y no condicionales')
# Usar bootstrapping para calcular intervalos de confianza:
Nboot = 1000
# Calcular indicadores de sensibilidad para Nboot remuestreos
KS_median, KS_mean, KS_max = PAWN.pawn_indices(X, E, n, Nboot=Nboot)
# KS_median, KS_mean y KS_max tienen tamano (Nboot, M)
# Calcular media e intervalos de confianza entre los resultados de los remuestreos:
KS_median_m, KS_median_lb, KS_median_ub = aggregate_boot(
KS_median) # shape (M,)
KS_mean_m, KS_mean_lb, KS_mean_ub = aggregate_boot(KS_mean) # shape (M,)
KS_max_m, KS_max_lb, KS_max_ub = aggregate_boot(KS_max) # shape (M,)
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_median_m_PAWN_OF.csv',
KS_median_m,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_median_lb_PAWN_OF.csv',
KS_median_lb,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_median_ub_PAWN_OF.csv',
KS_median_ub,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_mean_m_PAWN_OF.csv',
KS_mean_m,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_mean_lb_PAWN_OF.csv',
KS_mean_lb,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_mean_ub_PAWN_OF.csv',
KS_mean_ub,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_max_m_PAWN_OF.csv',
KS_max_m,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_max_lb_PAWN_OF.csv',
KS_max_lb,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_max_ub_PAWN_OF.csv',
KS_max_ub,
delimiter=',',
header='',
comments='')
# Plot bootstrapping results:
fig = plt.figure()
plt.subplot(131)
pf.boxplot1(KS_median_m,
S_lb=KS_median_lb,
S_ub=KS_median_ub,
X_Labels=X_Labels,
Y_Label='KS (median)')
plt.subplot(132)
pf.boxplot1(KS_mean_m,
S_lb=KS_mean_lb,
S_ub=KS_mean_ub,
X_Labels=X_Labels,
Y_Label='KS (mean)')
plt.subplot(133)
pf.boxplot1(KS_max_m,
S_lb=KS_max_lb,
S_ub=KS_max_ub,
X_Labels=X_Labels,
Y_Label='Ks (max)')
fig.savefig(r'%s\modeloFONAG\%s\sensibilidad' % (path, t_inputs[16]) +
r'\p_PAWN_OF_boot.png',
orientation='landscape',
papertype='a4')
print('Plot de sensibilidad PAWN con intervalos de confianza')
print(' ')
# %% Paso 4c: Analisis de sensibilidad PAWN para una variable dummy
print('Paso 4c: Analisis de sensibilidad PAWN para una variable dummy')
# Nombre de los parametros incluyendo el dummy:
X_Labels_dummy = X_Labels
X_Labels_dummy.append('dummy')
# Indicadores de sensibilidad usando bootstrap y la variable dummy:
# Usar bootstrapping para calcular intervalos de confianza:
Nboot = 1000
# Calcular los indicadores de sensibilidad usando los remuestreos.
# Se analiza solamente KS_max (y no KS_median y KS_mean) para mostrar resultados solamente.
_, _, KS_max, KS_dummy = PAWN.pawn_indices(X,
E,
n,
Nboot=Nboot,
dummy=True)
# KS_max has shape (Nboot, M), KS_dummy has shape (Nboot, )
# Cacular media e intervalos de confianza entre los remuestreos:
KS_m, KS_lb, KS_ub = aggregate_boot(np.column_stack((KS_max, KS_dummy)))
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_m_PAWN_OF.csv',
KS_m,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_lb_PAWN_OF.csv',
KS_lb,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\KS_ub_PAWN_OF.csv',
KS_ub,
delimiter=',',
header='',
comments='')
# Plot resultados de boostrap
fig = plt.figure() # plot main and total separately
pf.boxplot1(KS_m,
S_lb=KS_lb,
S_ub=KS_ub,
X_Labels=X_Labels_dummy,
Y_Label='KS')
fig.savefig(r'%s\modeloFONAG\%s\sensibilidad' % (path, t_inputs[16]) +
r'\p_PAWN_OF_dummy.png',
orientation='landscape',
papertype='a4')
print('Plot de sensibilidad PAWN con variable dummy')
elapsed_time = (time.time() - start) / 60
print(
'El proceso de analisis de sensibilidad PAWN para OF tomo %s minutos.'
% elapsed_time)
print('FIN PASO 4')
plt.show()
def rsa_OF(path, E_umbrales=1, cont=False):
print('CONVERGENCIA DEL ANALISIS DE SENSIBILIDAD REGIONAL PARA OF')
start = time.time()
print(' ')
# %% Paso 2a: Cargar datos del proceso de calibracion
print('Paso 2a: Cargar datos del proceso de calibracion')
inputs = open(r'%s\modeloFONAG\inputs\t_inputs.csv' % path, 'r')
t_inputs = inputs.read().split('\n')[1].split('\r')[0].split(',')
inputs.close()
print('Lectura de inputs')
# Crear carpeta de salidas
print('Crear carpeta de salidas')
os.system('md modeloFONAG\\%s\\sensibilidad' % t_inputs[16])
# Matriz de combinaciones de parametros
if cont:
X = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedXC.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
else:
X = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedX.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
print('Lectura de combinaciones de parametros')
# Nombres de parametros
X_Labels = open(
r'%s\modeloFONAG\%s\calibracion\X_Labels.csv' % (path, t_inputs[16]),
'r')
X_Labels = X_Labels.read().split('\n')[0].split(',')
X_Labels = map(str, X_Labels)
# Matriz de funciones objetivo
if cont:
E = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedEC.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
else:
E = genfromtxt('%s\modeloFONAG\%s\calibracion\sortedE.csv' %
(path, t_inputs[16]),
delimiter=',',
skip_header=1)
E = E[:, 5]
print('Lectura de funciones objetivo')
print(' ')
# %% Paso 2b: Analisis de sensibilidad regional
print('Paso 2b: Analisis de sensibilidad regional')
# Usar OF para RSA
# E_umbrales = 1
print('Definir umbrales de aceptacion')
# Aplicar Analisis de Sensibilidad Regional
E_max = max(E)
E_min = min(E)
if E_max <= E_umbrales:
print('Todas las simulaciones cumplen los umbrales')
elif E_min >= E_umbrales:
print('No existen simulaciones que cumplan los umbrales.')
else:
mvd, spread, irr, idxb = Rt.RSA_indices_thres(X, E, E_umbrales)
print('Aplicar Analisis de Sensibilidad Regional')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\idxb_RSA_OF.csv',
idxb,
delimiter=',',
header='',
comments='',
fmt='%s')
# Plot CDFs de los parametros
Rt.RSA_plot_thres(X,
idxb,
X_Labels=X_Labels,
str_legend=['cumplen', 'no cumplen'])
plt.savefig(r'%s\modeloFONAG\%s\sensibilidad' % (path, t_inputs[16]) +
r'\p_RSA_OF_cdf.png',
orientation='landscape',
papertype='a4')
print('Plot CDFs de los parametros')
# Calcular intervalos de confianza de los indicadores de sensibilidad usando bootstrapping
Nboot = 1000
mvd, spread, irr, idxb = Rt.RSA_indices_thres(X,
E,
E_umbrales,
Nboot=Nboot)
# mvd, spread e irr tienen tamano (Nboot, M)
print('Calcular intervalos de confianza')
# Calcular media e intervalos de confianza de MVD entre los remuestreos:
mvd_m, mvd_lb, mvd_ub = aggregate_boot(mvd) # shape (M,)
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\mvd_m_RSA_OF.csv',
mvd_m,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\mvd_lb_RSA_OF.csv',
mvd_lb,
delimiter=',',
header='',
comments='')
np.savetxt(r'%s\modeloFONAG\%s\calibracion' % (path, t_inputs[16]) +
r'\mvd_ub_RSA_OF.csv',
mvd_ub,
delimiter=',',
header='',
comments='')
# Plot results:
fig = plt.figure()
pf.boxplot1(mvd_m,
X_Labels=X_Labels,
Y_Label='Sensibilidad MVD',
S_lb=mvd_lb,
S_ub=mvd_ub)
fig.savefig(r'%s\modeloFONAG\%s\sensibilidad' % (path, t_inputs[16]) +
r'\p_MVD_OF_boot.png',
orientation='landscape',
papertype='a4')
print('Plot de sensibilidad MVD con intervalos de confianza')
elapsed_time = (time.time() - start) / 60
print(
'El proceso de analisis de sensibilidad RSA para OF tomo %s minutos.' %
elapsed_time)
print('FIN PASO 2')
plt.show()