def extract_csr():
directory = r"C:\Users\FEli\OneDrive - Golder Associates\Projects\2020\PRDP\mean_profile_n0.15\Output\\"
dir_list = os.listdir(directory)
for cur_dir in dir_list:
motions = os.listdir(directory + cur_dir + '/profile_0/')
depth = np.zeros(0)
results_list = []
title = 'depth'
for motion_ii, motion in enumerate(motions):
print(directory + cur_dir + '/' + motion + '/' + 'deepsoilout.db3')
print("debug1")
print(motion_ii)
ds = deepsoil_plotter(directory + cur_dir + '/profile_0/' +
motion + '/' + 'deepsoilout.db3')
depth = ds.get_elem_mid_depth().squeeze()
csr = ds.get_max_stress_ratio().squeeze()
if motion_ii == 0:
results_list.append(depth)
title += ',' + motion
results_list.append(csr.squeeze())
to_write = np.stack(results_list, axis=1)
with open(directory + cur_dir + '.csv', 'w') as f:
f.write(title + '\n')
with open(directory + cur_dir + '.csv', 'ba') as f:
np.savetxt(f, to_write, delimiter=',')
def print_profiles():
directory = r"C:\Users\FEli\OneDrive - Golder Associates\Projects\2020\PRDP\mean_profile_n0.15\Output\\"
dir_list = os.listdir(directory)
for cur_dir in dir_list:
motions = os.listdir(directory + cur_dir + '/profile_0/')
depth = np.zeros(0)
fig = plt.figure(figsize=(8, 6), dpi=200, tight_layout=True)
ax_pgd = fig.add_subplot(1, 4, 1)
ax_pga = fig.add_subplot(1, 4, 2)
ax_gmx = fig.add_subplot(1, 4, 3)
ax_csr = fig.add_subplot(1, 4, 4)
num_motions = len(motions)
for motion_ii, motion in enumerate(motions):
ds = deepsoil_plotter(directory + cur_dir + '/profile_0/' +
motion + '/' + 'deepsoilout.db3')
depth = ds.get_elem_mid_depth().squeeze()
pgd = ds.get_relPGD_profile().squeeze()
pga = ds.get_totPGA_profile().squeeze()
gmx = ds.get_max_strain().squeeze()
csr = ds.get_max_stress_ratio().squeeze()
if motion_ii == 0:
pgd_avg = pgd / num_motions
pga_avg = pga / num_motions
gmx_avg = gmx / num_motions
csr_avg = csr / num_motions
else:
pgd_avg += pgd / num_motions
pga_avg += pga / num_motions
gmx_avg += gmx / num_motions
csr_avg += csr / num_motions
ax_pgd.plot(pgd, depth, 'r-', lw=1.0, alpha=0.25)
ax_pga.plot(pga, depth, 'r-', lw=1.0, alpha=0.25)
ax_gmx.plot(gmx, depth, 'r-', lw=1.0, alpha=0.25)
ax_csr.plot(csr, depth, 'r-', lw=1.0, alpha=0.25)
ax_pgd.plot(pgd_avg, depth, 'r-', lw=2.0)
ax_pga.plot(pga_avg, depth, 'r-', lw=2.0)
ax_gmx.plot(gmx_avg, depth, 'r-', lw=2.0)
ax_csr.plot(csr_avg, depth, 'r-', lw=2.0)
decorate_plot(ax_pgd)
ax_pgd.set_xlabel("PGD (m)")
ax_pgd.set_ylabel("Depth (m)")
decorate_plot(ax_pga)
ax_pga.set_xlabel("PGA (g)")
decorate_plot(ax_gmx)
ax_gmx.set_xlabel("$\gamma_{max}$ (%)")
ax_gmx.set_xlim(right=1.0)
decorate_plot(ax_csr)
ax_csr.set_xlabel("CSR")
fig.savefig(directory + cur_dir + '_profile.png')
def run_parallel_aux(cwd):
# print(os.path.join(cwd, 'deepsoilout.db3'))
temp_file = os.path.join(cwd, 'deepsoilout.db3')
if os.path.isfile(temp_file):
ds = deepsoil_plotter(os.path.join(cwd, 'deepsoilout.db3'))
depth = ds.get_elem_mid_depth().squeeze()
pgd = ds.get_relPGD_profile().squeeze()
pga = ds.get_totPGA_profile().squeeze()
gmx = ds.get_max_strain().squeeze()
csr = 0.65 * ds.get_max_stress_ratio().squeeze()
with open(os.path.join(cwd, 'deepsoilin.txt'), 'r') as f:
ds_input = f.read().replace('\n',' ').replace('\t',' ').split()
Vs = []
unit_weight = []
in_depth = []
cur_depth = 0.0
layer_num = 0
water_table = 1
pwp = []
tot_stress = []
cur_pwp = 0.0
cur_tot_stress = 0.0
eff_stress = []
for data in ds_input:
if data.startswith('[WATER_TABLE]'):
gwt = int(data.split(':')[1][1:-1])
if data.startswith('[LAYER]'):
if 'TOP' in data:
break
layer_num = int(data.split(':')[1][1:-1])
if data.startswith('[THICKNESS]'):
thickness = float(data.split(':')[1][1:-1])
in_depth.append(cur_depth + thickness / 2.0)
if data.startswith('[WEIGHT]'):
unit_weight.append(float(data.split(':')[1][1:-1]))
cur_tot_stress += unit_weight[-1] * thickness / 2.0
if layer_num > gwt:
cur_pwp += 9.81 * thickness / 2.0
tot_stress.append(cur_tot_stress)
pwp.append(cur_pwp)
eff_stress.append(cur_tot_stress-cur_pwp)
cur_tot_stress += unit_weight[-1] * thickness / 2.0
if layer_num > gwt:
cur_pwp += 9.81 * thickness / 2.0
if data.startswith('[SHEAR]'):
Vs.append(float(data.split(':')[1][1:-1]))
tot_stress = np.array(tot_stress)
eff_stress = np.array(eff_stress)
Vs = np.array(Vs)
fs, pl, _, _ = liquefaction_triggering_Kayen(depth, Vs, csr=csr, sig_tot=tot_stress, sig_eff=eff_stress, Mw=7.7, fc=0.0*Vs)
fs_simpl, pl_simpl, csr_simpl, crr_simpl = liquefaction_triggering_Kayen(depth, Vs, sig_tot=tot_stress, sig_eff=eff_stress, pga=0.24, Mw=7.7, fc=0.0*Vs)
data_to_write = np.stack((depth, Vs, eff_stress, tot_stress, pgd, pga, gmx, csr, fs, pl, csr_simpl, crr_simpl, fs_simpl, pl_simpl), axis=1)
with open(os.path.join(cwd, 'results.csv'), 'w') as f:
f.write('depth, Vs, eff_stress, tot_stress, pgd, pga, gamma_max, csr, fs, pl, csr_simpl, crr_simpl, fs_simpl, pl_simpl\n')
with open(os.path.join(cwd, 'results.csv'), 'ab') as f:
np.savetxt(f, data_to_write, delimiter=',')
del ds
os.remove(os.path.join(cwd, 'deepsoilout.db3'))
os.remove(os.path.join(cwd, 'deepsoilout_el.db3'))
os.remove(os.path.join(cwd, 'strain.txt'))
print(f'Done with {cwd}')
def plot_profiles_fs_excel():
import openpyxl as xl
directory = r"C:\Users\FEli\OneDrive - Golder Associates\Projects\2020\PRDP\mean_profile_n0.15\Output\\"
profiles = [
# 'Bhb-17-Simplified_new',
# # 'Bhd-01-Simplified',
# 'Bhr-01-Simplified_new',
# 'Bhr-04-Simplified_new',
# 'Bhd-01-Simplified_new'
'Darendeli',
'Zhang'
]
motions = [
'Motion_Matched_Darfiel_N80E',
'Motion_Matched_Duzce_531_N',
'Motion_Matched_FKS_EW',
'Motion_Matched_Hokkaid_63_BL',
'Motion_Matched_Ibaraki_90',
'Motion_matched_LANDERS_DSP090',
'Motion_Matched_Manjil_L',
'Motion_Matched_Romania_NE',
'Motion_Matched_Sitka Alaska_90',
'Motion_Matched_SOUTHERN_SUMATRA_N_BL',
]
num_motions = len(motions)
fig = plt.figure(figsize=(16, 9), dpi=200)
ax_pgd = fig.add_subplot(1, 7, 1)
ax_pga = fig.add_subplot(1, 7, 2)
ax_gmx = fig.add_subplot(1, 7, 3)
ax_csr = fig.add_subplot(1, 7, 4)
ax_fs = fig.add_subplot(1, 7, 5)
ax_pl = fig.add_subplot(1, 7, 6)
ax_vs = fig.add_subplot(1, 7, 7)
# wb = xl.load_workbook(r"C:\Users\FEli\OneDrive - Golder Associates\Projects\2020\PRDP\mean_profile_n0.15\%s.xlsx"%profile, data_only=True)
wb = xl.load_workbook(
r"c:\Users\FEli\Desktop\PRDP\Bhr-04\Bhr-04-Darendeli_soft2.xlsx",
data_only=True)
sh = wb["Info"]
waterlevel = float(sh[f'D7'].value)
sh = wb["Deepsoil-Calibrated"]
sublayer = 1
Vs = []
unit_weight = []
in_depth = []
pwp = []
tot_stress = []
eff_stress = []
fc = []
cur_depth = 0
soil = []
while True:
if sh[f'A{sublayer + 1}'].value == None:
break
cur_depth = cur_depth + float(sh[f'C{sublayer + 1}'].value) / 2.0
in_depth.append(cur_depth)
unit_weight.append(float(sh[f'E{sublayer + 1}'].value))
pwp.append(float(sh[f'R{sublayer + 1}'].value))
tot_stress.append(float(sh[f'Q{sublayer + 1}'].value))
eff_stress.append(float(sh[f'S{sublayer + 1}'].value))
fc.append(float(sh[f'U{sublayer+1}'].value))
Vs.append(float(sh[f'F{sublayer + 1}'].value))
soil.append(str(sh[f'B{sublayer+1}'].value))
cur_depth = cur_depth + float(sh[f'C{sublayer + 1}'].value) / 2.0
sublayer += 1
tot_stress = np.array(tot_stress)
eff_stress = np.array(eff_stress)
Vs = np.array(Vs)
fc = np.array(fc)
csr_avg_all = 0
pga_surface_avg = 0
for profile in profiles:
depth = np.zeros(0)
for motion_ii, motion in enumerate(motions):
# ds = deepsoil_plotter(os.path.join(directory,profile ,motion, 'deepsoilout.db3'))
ds = deepsoil_plotter(
os.path.join(directory, profile, 'profile_0', motion,
'deepsoilout.db3'))
depth = ds.get_elem_mid_depth().squeeze()
pgd = ds.get_relPGD_profile().squeeze()
pga = ds.get_totPGA_profile().squeeze()
gmx = ds.get_max_strain().squeeze()
csr = ds.get_max_stress_ratio().squeeze() * 0.65
pga_surface = pga[0]
fs, pl, _, _ = liquefaction_triggering_Kayen(depth,
Vs,
csr=csr,
sig_tot=tot_stress,
sig_eff=eff_stress,
Mw=8.2,
fc=fc)
# fs_simpl, pl_simpl, csr_simpl, crr_simpl = liquefaction_triggering_Kayen(depth, Vs, sig_tot=tot_stress, sig_eff=eff_stress, pga=0.24, Mw=8.2, fc=fc)
fs_simpl, pl_simpl, csr_simpl, crr_simpl = liquefaction_triggering_Kayen(
depth,
Vs,
sig_tot=tot_stress,
sig_eff=eff_stress,
pga=pga_surface,
Mw=8.2,
fc=fc)
csr_avg_all += csr / num_motions / 2
pga_surface_avg += pga_surface / num_motions / 2
if motion_ii == 0:
pgd_avg = pgd / num_motions
pga_avg = pga / num_motions
gmx_avg = gmx / num_motions
csr_avg = csr / num_motions
fs_avg = fs / num_motions
pl_avg = pl / num_motions
csr_simpl_avg = csr_simpl / num_motions
fs_simpl_avg = fs_simpl / num_motions
pl_simpl_avg = pl_simpl / num_motions
else:
pgd_avg += pgd / num_motions
pga_avg += pga / num_motions
gmx_avg += gmx / num_motions
csr_avg += csr / num_motions
fs_avg += fs / num_motions
pl_avg += pl / num_motions
csr_simpl_avg += csr_simpl / num_motions
fs_simpl_avg += fs_simpl / num_motions
pl_simpl_avg += pl_simpl / num_motions
ax_pgd.plot(pgd, depth, 'k-', lw=1.0, alpha=0.25)
ax_pga.plot(pga, depth, 'k-', lw=1.0, alpha=0.25)
ax_gmx.plot(gmx, depth, 'k-', lw=1.0, alpha=0.25)
ax_csr.plot(csr, depth, 'k-', lw=1.0, alpha=0.25)
ax_csr.plot(csr_simpl, depth, 'b-', lw=1.0, alpha=0.25)
fs[depth < waterlevel] = np.NAN
fs_simpl[depth < waterlevel] = np.NAN
pl[depth < waterlevel] = np.NAN
pl_simpl[depth < waterlevel] = np.NAN
ax_fs.plot(fs, depth, 'k-', lw=1.0, alpha=0.25)
ax_fs.plot(fs_simpl, depth, 'b-', lw=1.0, alpha=0.25)
ax_pl.plot(pl, depth, 'k-', lw=1.0, alpha=0.25)
ax_pl.plot(pl_simpl, depth, 'b-', lw=1.0, alpha=0.25)
ax_pgd.plot(pgd_avg, depth, 'k-', lw=2.0)
ax_pga.plot(pga_avg, depth, 'k-', lw=2.0)
ax_gmx.plot(gmx_avg, depth, 'k-', lw=2.0)
ax_csr.plot(csr_avg, depth, 'k-', lw=2.0)
ax_csr.plot(csr_simpl_avg, depth, 'b-', lw=2.0)
fs_avg[depth < waterlevel] = np.NAN
fs_simpl_avg[depth < waterlevel] = np.NAN
pl_avg[depth < waterlevel] = np.NAN
pl_simpl_avg[depth < waterlevel] = np.NAN
ax_fs.plot(fs_avg, depth, 'k-', lw=2.0)
ax_fs.plot(fs_simpl_avg, depth, 'b-', lw=2.0)
ax_pl.plot(pl_avg, depth, 'k-', lw=2.0)
ax_pl.plot(pl_simpl_avg, depth, 'b-', lw=2.0)
ax_vs.plot(Vs, depth, 'k-', lw=2.0)
Vs_test = []
ii = 0
for idepth in depth:
if (soil[ii].startswith('Compacted')):
Vs_test.append(0.018 * idepth**3.0 - 0.89 * idepth**2.0 +
16.2 * idepth + 104.0)
elif (soil[ii].startswith('Stiff')):
Vs_test.append(300)
elif (soil[ii].startswith('Soft')):
Vs_test.append(1.5 * idepth + 117.5)
else:
Vs_test.append(np.NAN)
ii += 1
ax_vs.plot(Vs_test, depth, 'g-', lw=2.0, label="JD-2018")
lines = [
matplotlib.lines.Line2D([], [],
color='k',
linestyle='-',
lw=2.0,
label='Site-Specific'),
matplotlib.lines.Line2D([], [],
color='b',
linestyle='-',
lw=2.0,
label='Simplified'),
]
ax_pl.legend(handles=lines, loc='lower right')
## plot the average CSR
fs, pl, _, _ = liquefaction_triggering_Kayen(depth,
Vs,
csr=csr_avg_all,
sig_tot=tot_stress,
sig_eff=eff_stress,
Mw=8.2,
fc=fc)
fs_simpl, pl_simpl, csr_simpl, crr_simpl = liquefaction_triggering_Kayen(
depth,
Vs,
sig_tot=tot_stress,
sig_eff=eff_stress,
pga=pga_surface_avg,
Mw=8.2,
fc=fc)
fs[depth < waterlevel] = np.NAN
pl[depth < waterlevel] = np.NAN
ax_fs.plot(fs, depth, 'r-', lw=2.0)
ax_pl.plot(pl, depth, 'r-', lw=2.0)
ax_fs.text(0, waterlevel - 0.5, "Water Level")
ax_fs.plot([0, 2], [waterlevel, waterlevel], 'k-.', lw=2.0)
ax_pl.text(0, waterlevel - 0.5, "Water Level")
ax_pl.plot([0, 1], [waterlevel, waterlevel], 'k-.', lw=2.0)
fs_simpl[depth < waterlevel] = np.NAN
pl_simpl[depth < waterlevel] = np.NAN
ax_fs.plot(fs_simpl, depth, 'b--', lw=2.0)
ax_pl.plot(pl_simpl, depth, 'b--', lw=2.0)
decorate_plot(ax_pgd)
ax_pgd.set_xlabel("PGD (m)")
ax_pgd.set_ylabel("Depth (m)")
decorate_plot(ax_pga)
ax_pga.axvline(x=0.24, color='b', lw=2.0)
ax_pga.set_xlabel("PGA (g)")
decorate_plot(ax_gmx)
ax_gmx.set_xlabel("$\gamma_{max}$ (%)")
ax_gmx.set_xlim(right=1.0)
decorate_plot(ax_csr)
ax_csr.set_xlabel("CSR")
decorate_plot(ax_fs)
ax_fs.set_xlabel("FS")
ax_fs.set_xlim(0.0, 2.0)
ax_fs.axvline(x=1.3, color='r', linestyle='--')
decorate_plot(ax_pl)
ax_pl.set_xlabel("PL")
decorate_plot(ax_vs)
ax_vs.set_xlabel("Vs")
fig.suptitle(profile)
fig.tight_layout(rect=[0.05, 0.05, 0.95, 0.95])
fig.savefig(os.path.join(directory, profile + '.png'))
def plot_profiles_fs():
directory = r"C:\Users\FEli\OneDrive - Golder Associates\Projects\2020\PRDP\mean_profile_n0.15\Output\\"
profiles = [
# 'Bhb-17-Simplified',
# 'Bhd-01-Simplified',
# 'Bhr-01-Simplified',
# 'Bhr-04-Simplified',
# 'Bhd-01-Simplified_new'
'Bhr-04-Darendeli_soft2'
]
motions = [
'Matched_Darfiel_N80E',
'Matched_Duzce_531_N',
'Matched_FKS_EW',
'Matched_Hokkaid_63_BL',
'Matched_Ibaraki_90',
'matched_LANDERS_DSP090',
'Matched_Manjil_L',
'Matched_Romania_NE',
'Matched_Sitka Alaska_90',
'Matched_SOUTHERN_SUMATRA_N_BL',
]
num_motions = len(motions)
for profile in profiles:
depth = np.zeros(0)
fig = plt.figure(figsize=(16, 9), dpi=200)
ax_pgd = fig.add_subplot(1, 6, 1)
ax_pga = fig.add_subplot(1, 6, 2)
ax_gmx = fig.add_subplot(1, 6, 3)
ax_csr = fig.add_subplot(1, 6, 4)
ax_fs = fig.add_subplot(1, 6, 5)
ax_pl = fig.add_subplot(1, 6, 6)
for motion_ii, motion in enumerate(motions):
ds = deepsoil_plotter(
os.path.join(directory, profile, motion, 'deepsoilout.db3'))
with open(
os.path.join(directory, profile, motion, 'deepsoilin.txt'),
'r') as f:
ds_input = f.read().replace('\n', ' ').replace('\t',
' ').split()
depth = ds.get_elem_mid_depth().squeeze()
pgd = ds.get_relPGD_profile().squeeze()
pga = ds.get_totPGA_profile().squeeze()
gmx = ds.get_max_strain().squeeze()
csr = ds.get_max_stress_ratio().squeeze() * 0.65
Vs = []
unit_weight = []
in_depth = []
cur_depth = 0.0
layer_num = 0
water_table = 1
pwp = []
tot_stress = []
cur_pwp = 0.0
cur_tot_stress = 0.0
eff_stress = []
for data in ds_input:
if data.startswith('[WATER_TABLE]'):
gwt = int(data.split(':')[1][1:-1])
if data.startswith('[LAYER]'):
if 'TOP' in data:
break
layer_num = int(data.split(':')[1][1:-1])
if data.startswith('[THICKNESS]'):
thickness = float(data.split(':')[1][1:-1])
in_depth.append(cur_depth + thickness / 2.0)
if data.startswith('[WEIGHT]'):
unit_weight.append(float(data.split(':')[1][1:-1]))
cur_tot_stress += unit_weight[-1] * thickness / 2.0
if layer_num > gwt:
cur_pwp += 9.81 * thickness / 2.0
tot_stress.append(cur_tot_stress)
pwp.append(cur_pwp)
eff_stress.append(cur_tot_stress - cur_pwp)
cur_tot_stress += unit_weight[-1] * thickness / 2.0
if layer_num > gwt:
cur_pwp += 9.81 * thickness / 2.0
if data.startswith('[SHEAR]'):
Vs.append(float(data.split(':')[1][1:-1]))
tot_stress = np.array(tot_stress)
eff_stress = np.array(eff_stress)
Vs = np.array(Vs)
fs, pl, _, _ = liquefaction_triggering_Kayen(depth,
Vs,
csr=csr,
sig_tot=tot_stress,
sig_eff=eff_stress,
Mw=7.7,
fc=0.0 * Vs)
fs_simpl, pl_simpl, csr_simpl, crr_simpl = liquefaction_triggering_Kayen(
depth,
Vs,
sig_tot=tot_stress,
sig_eff=eff_stress,
pga=0.24,
Mw=7.7,
fc=0.0 * Vs)
if motion_ii == 0:
pgd_avg = pgd / num_motions
pga_avg = pga / num_motions
gmx_avg = gmx / num_motions
csr_avg = csr / num_motions
fs_avg = fs / num_motions
pl_avg = pl / num_motions
csr_simpl_avg = csr_simpl / num_motions
fs_simpl_avg = fs_simpl / num_motions
pl_simpl_avg = pl_simpl / num_motions
else:
pgd_avg += pgd / num_motions
pga_avg += pga / num_motions
gmx_avg += gmx / num_motions
csr_avg += csr / num_motions
fs_avg += fs / num_motions
pl_avg += pl / num_motions
csr_simpl_avg += csr_simpl / num_motions
fs_simpl_avg += fs_simpl / num_motions
pl_simpl_avg += pl_simpl / num_motions
ax_pgd.plot(pgd, depth, 'k-', lw=1.0, alpha=0.25)
ax_pga.plot(pga, depth, 'k-', lw=1.0, alpha=0.25)
ax_gmx.plot(gmx, depth, 'k-', lw=1.0, alpha=0.25)
ax_csr.plot(csr, depth, 'k-', lw=1.0, alpha=0.25)
ax_csr.plot(csr_simpl, depth, 'b-', lw=1.0, alpha=0.25)
ax_fs.plot(fs, depth, 'k-', lw=1.0, alpha=0.25)
ax_fs.plot(fs_simpl, depth, 'b-', lw=1.0, alpha=0.25)
ax_pl.plot(pl, depth, 'k-', lw=1.0, alpha=0.25)
ax_pl.plot(pl_simpl, depth, 'b-', lw=1.0, alpha=0.25)
ax_pgd.plot(pgd_avg, depth, 'k-', lw=2.0)
ax_pga.plot(pga_avg, depth, 'k-', lw=2.0)
ax_gmx.plot(gmx_avg, depth, 'k-', lw=2.0)
ax_csr.plot(csr_avg, depth, 'k-', lw=2.0)
ax_csr.plot(csr_simpl_avg, depth, 'b-', lw=2.0)
ax_fs.plot(fs_avg, depth, 'k-', lw=2.0)
ax_fs.plot(fs_simpl_avg, depth, 'b-', lw=2.0)
ax_pl.plot(pl_avg, depth, 'k-', lw=2.0)
ax_pl.plot(pl_simpl_avg, depth, 'b-', lw=2.0)
decorate_plot(ax_pgd)
ax_pgd.set_xlabel("PGD (m)")
ax_pgd.set_ylabel("Depth (m)")
decorate_plot(ax_pga)
ax_pga.axvline(x=0.24, color='b', lw=2.0)
ax_pga.set_xlabel("PGA (g)")
decorate_plot(ax_gmx)
ax_gmx.set_xlabel("$\gamma_{max}$ (%)")
ax_gmx.set_xlim(right=1.0)
decorate_plot(ax_csr)
ax_csr.set_xlabel("CSR")
decorate_plot(ax_fs)
ax_fs.set_xlabel("FS")
ax_fs.set_xlim(0.0, 2.0)
ax_fs.axvline(x=1.3, color='r', linestyle='--')
decorate_plot(ax_pl)
ax_pl.set_xlabel("PL")
lines = [
matplotlib.lines.Line2D([], [],
color='k',
linestyle='-',
lw=2.0,
label='Site-Specific'),
matplotlib.lines.Line2D([], [],
color='b',
linestyle='-',
lw=2.0,
label='Simplified'),
]
ax_pl.legend(handles=lines, loc='lower right')
fig.suptitle(profile)
fig.tight_layout(rect=[0.05, 0.05, 0.95, 0.95])
fig.savefig(os.path.join(directory, profile + '.png'))