ss_a1_model.save(
os.path.join(path_model_cell_out +
str('/ss_a1_model_metabolic_signaling')))
ss_a2_model.save(
os.path.join(path_model_cell_out +
str('/ss_a2_model_metabolic_signaling')))
ss_b1_model.save(
os.path.join(path_model_cell_out +
str('/ss_b1_model_metabolic_signaling')))
ss_b2_model.save(
os.path.join(path_model_cell_out +
str('/ss_b2_model_metabolic_signaling')))
print('MODELS EXPORTED to "{}"'.format(path_model_cell_out))
ss_dense_hp = tfm_kt.def_hp(ss_dense_hp, 'ss_dense_hp')
ss_a1_hp = tfm_kt.def_hp(ss_a1_hp, 'ss_a1_hp')
ss_a2_hp = tfm_kt.def_hp(ss_a2_hp, 'ss_a2_hp')
ss_b1_hp = tfm_kt.def_hp(ss_b1_hp, 'ss_b1_hp')
ss_b2_hp = tfm_kt.def_hp(ss_b2_hp, 'ss_b2_hp')
df_hp = pd.concat([ss_dense_hp, ss_a1_hp, ss_a2_hp, ss_b1_hp, ss_b2_hp])
df_hp = df_hp.set_index('hp')
df_hp.to_csv(os.path.join(path_output_result +
'/kt_hyperparameters_metabolic_signaling_' +
n_out_cell_type + '.txt'),
sep=';')
print('RESULT EXPORTED to "{}"'.format(path_output_result))
project_name_='kt_ss_p2_metabolic_signaling_experiment_',
second_layer=False,
path_=path_hyperband_,
epochs_=epochs_default,
batch_size_=batch_size_default).build()
ss_p1_model.save(os.path.join(path_model + str('/ss_p1_model_default')))
ss_p2_sig_model.save(
os.path.join(path_model + str('/ss_p2_model_signaling')))
ss_p2_met_sig_model.save(
os.path.join(path_model + str('/ss_p2_model_metabolic_signaling')))
print('MODELS EXPORTED to "{}"'.format(path_model))
ss_p1_hp = tfm_kt.def_hp(ss_p1_hp, 'ss_p1_hp')
ss_p2_sig_hp = tfm_kt.def_hp(ss_p2_sig_hp, 'ss_p2_sig_hp')
ss_p2_met_sig_hp = tfm_kt.def_hp(ss_p2_met_sig_hp, 'ss_p2_met_sig_hp')
df_hp = pd.concat([ss_p1_hp, ss_p2_sig_hp, ss_p2_met_sig_hp])
df_hp = df_hp.set_index('hp')
df_hp.to_csv(os.path.join(path_output_result +
'/kt_hyperparameters_design_p_no_co_' +
str(i_scaling) + '.txt'),
sep=';')
print('RESULT EXPORTED to "{}"'.format(path_output_result))
print("\n\nDesign P dense\n")
print(ss_p1_model.summary())