Skip to content

Commit

Permalink
update
Browse files Browse the repository at this point in the history
  • Loading branch information
@CB-quakemodel
CB-quakemodel committed Jul 7, 2024
1 parent 52c9bd1 commit 99d7a00
Showing 1 changed file with 32 additions and 105 deletions.
137 changes: 32 additions & 105 deletions openquake/smt/comparison/utils_compare_gmpes.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -277,9 +277,10 @@ def plot_spectra_util(config, output_directory, obs_spectra):
linewidth=0.75, linestyle='-.')

# Weight the predictions using logic tree weights
gmc_vals = spectra_data(
gmpe, config.Nstd, rs_50p, rs_plus_sigma, rs_minus_sigma,
lt_vals, lt_vals_plus, lt_vals_minus, config)
gmc_vals = spectra_data(gmpe,
rs_50p, rs_plus_sigma, rs_minus_sigma,
lt_vals, lt_vals_plus, lt_vals_minus,
config)

# Plot obs spectra if required
if obs_spectra is not None:
Expand Down Expand Up @@ -876,117 +877,43 @@ def _get_imts(max_period):
return imt_list, periods


def spectra_data(gmpe, Nstd, rs_50p, rs_plus_sigma, rs_minus_sigma,
def spectra_data(gmpe, rs_50p, rs_plus_sigma, rs_minus_sigma,
lt_vals, lt_vals_plus, lt_vals_minus, config):
"""
If required get the logic tree weighted predictions
"""
# Logic tree #1
if config.lt_weights_gmc1 == None:
pass
elif gmpe in config.lt_weights_gmc1:
if config.lt_weights_gmc1[gmpe] != None:
rs_50p_weighted_gmc1 = {}
rs_plus_sigma_weighted_gmc1 = {}
rs_minus_sigma_weighted_gmc1 = {}
for idx, rs in enumerate(rs_50p):
rs_50p_weighted_gmc1[idx] = rs_50p[idx]*config.lt_weights_gmc1[gmpe]
if Nstd != 0:
rs_plus_sigma_weighted_gmc1[idx] = rs_plus_sigma[
idx]*config.lt_weights_gmc1[gmpe]
rs_minus_sigma_weighted_gmc1[idx] = rs_minus_sigma[
idx]*config.lt_weights_gmc1[gmpe]

# If present store the weighted median for the GMPE
lt_vals[0][gmpe] = {'median': rs_50p_weighted_gmc1}

# And if Nstd != 0 store these weighted branches too
if Nstd != 0:
lt_vals_plus[0][gmpe,'p_sig'] = {
'plus_sigma': rs_plus_sigma_weighted_gmc1}
lt_vals_minus[0][gmpe,'m_sig'] = {
'minus_sigma': rs_minus_sigma_weighted_gmc1}

# Logic tree #2
if config.lt_weights_gmc2 == None:
pass
elif gmpe in config.lt_weights_gmc2:
if config.lt_weights_gmc2[gmpe] != None:
rs_50p_weighted_gmc2 = {}
rs_plus_sigma_weighted_gmc2, rs_minus_sigma_weighted_gmc2 = {}, {}
for idx, rs in enumerate(rs_50p):
rs_50p_weighted_gmc2[idx] = rs_50p[idx]*config.lt_weights_gmc2[gmpe]
if Nstd != 0:
rs_plus_sigma_weighted_gmc2[idx] = rs_plus_sigma[
idx]*config.lt_weights_gmc2[gmpe]
rs_minus_sigma_weighted_gmc2[idx] = rs_minus_sigma[
idx]*config.lt_weights_gmc2[gmpe]

# If present store the weighted median for the GMPE
lt_vals[1][gmpe] = {'median': rs_50p_weighted_gmc2}

# And if Nstd != 0 store these weighted branches too
if Nstd != 0:
lt_vals_plus[1][gmpe,'p_sig'] = {
'plus_sigma': rs_plus_sigma_weighted_gmc2}
lt_vals_minus[1][gmpe,'m_sig'] = {
'minus_sigma': rs_minus_sigma_weighted_gmc2}

# Logic tree #3
if config.lt_weights_gmc3 == None:
pass
elif gmpe in config.lt_weights_gmc3:
if config.lt_weights_gmc3[gmpe] != None:
rs_50p_weighted_gmc3 = {}
rs_plus_sigma_weighted_gmc3, rs_minus_sigma_weighted_gmc3 = {}, {}
for idx, rs in enumerate(rs_50p):
rs_50p_weighted_gmc3[idx] = rs_50p[idx]*config.lt_weights_gmc3[gmpe]
if Nstd != 0:
rs_plus_sigma_weighted_gmc3[idx] = rs_plus_sigma[
idx]*config.lt_weights_gmc3[gmpe]
rs_minus_sigma_weighted_gmc3[idx] = rs_minus_sigma[
idx]*config.lt_weights_gmc3[gmpe]

# If present store the weighted median for the GMPE
lt_vals[2][gmpe] = {'median': rs_50p_weighted_gmc3}

# And if Nstd != 0 store these weighted branches too
if Nstd != 0:
lt_vals_plus[2][gmpe,'p_sig'] = {
'plus_sigma': rs_plus_sigma_weighted_gmc3}
lt_vals_minus[2][gmpe,'m_sig'] = {
'minus_sigma': rs_minus_sigma_weighted_gmc3}

# Logic tree #4
if config.lt_weights_gmc4 == None:
pass
elif gmpe in config.lt_weights_gmc4:
if config.lt_weights_gmc4[gmpe] != None:
rs_50p_weighted_gmc4 = {}
rs_plus_sigma_weighted_gmc4, rs_minus_sigma_weighted_gmc4 = {}, {}
for idx, rs in enumerate(rs_50p):
rs_50p_weighted_gmc4[idx] = rs_50p[idx]*config.lt_weights_gmc4[gmpe]
if Nstd != 0:
rs_plus_sigma_weighted_gmc4[idx] = rs_plus_sigma[
idx]*config.lt_weights_gmc4[gmpe]
rs_minus_sigma_weighted_gmc4[idx] = rs_minus_sigma[
idx]*config.lt_weights_gmc4[gmpe]

# If present store the weighted median for the GMPE
lt_vals[3][gmpe] = {'median': rs_50p_weighted_gmc4}

# And if Nstd != 0 store these weighted branches too
if Nstd != 0:
lt_vals_plus[3][gmpe,'p_sig'] = {
'plus_sigma': rs_plus_sigma_weighted_gmc4}
lt_vals_minus[3][gmpe,'m_sig'] = {
'minus_sigma': rs_minus_sigma_weighted_gmc4}
gmc_weights = [config.lt_weights_gmc1, config.lt_weights_gmc2,
config.lt_weights_gmc3, config.lt_weights_gmc4]
for idx_gmc, gmc in enumerate(gmc_weights):
if gmc_weights[idx_gmc] is None:
pass
elif gmpe in gmc_weights[idx_gmc]:
if gmc_weights[idx_gmc][gmpe] is not None:
rs_50p_w, rs_plus_sigma_w, rs_minus_sigma_w = {}, {}, {}
for idx, rs in enumerate(rs_50p):
rs_50p_w[idx] = rs_50p[idx]*gmc_weights[
idx_gmc][gmpe]
if config.Nstd != 0:
rs_plus_sigma_w[idx] = rs_plus_sigma[
idx]*gmc_weights[idx_gmc][gmpe]
rs_minus_sigma_w[idx] = rs_minus_sigma[
idx]*gmc_weights[idx_gmc][gmpe]

# Store the weighted median for the GMPE
lt_vals[idx_gmc][gmpe] = {'median': rs_50p_w}

# And if Nstd != 0 store these weighted branches too
if config.Nstd != 0:
lt_vals_plus[idx_gmc][gmpe,'p_sig'] = {
'plus_sigma': rs_plus_sigma_w}
lt_vals_minus[idx_gmc][gmpe,'m_sig'] = {
'minus_sigma': rs_minus_sigma_w}

gmc1_vals = [lt_vals[0], lt_vals_plus[0], lt_vals_minus[0]]
gmc2_vals = [lt_vals[1], lt_vals_plus[1], lt_vals_minus[1]]
gmc3_vals = [lt_vals[2], lt_vals_plus[2], lt_vals_minus[2]]
gmc4_vals = [lt_vals[3], lt_vals_plus[3], lt_vals_minus[3]]

return gmc1_vals, gmc2_vals, gmc3_vals, gmc4_vals


Expand Down

0 comments on commit 99d7a00

Please sign in to comment.