def test_plot_steps(self):
figs, AX = ta.plot_steps(traces,
skip=[],
palette={},
legend_args={},
plot_args={'marker': '+', 'linewidth':0}
)
assert figs
assert AX
's': The scipy return value if applicable
'f': The full trace including calls to local optimizers if applicable
'''
accepted = result['a']
plot_index = {1: ['OD600', 'Glu', 'Fluor', specific_growth]}
titles = {1: dict(zip(plot_index[1], plot_index[1]))}
#Create our own Figure and Axes objects for cutom layouts
fig = plt.figure()
AX_ = [fig.add_subplot(2, 2, i + 1) for i in range(4)]
AX = {
1: {
'OD600': AX_[0],
'Fluor': AX_[1],
'Glu': AX_[2],
specific_growth: AX_[3]
}
}
cf.plot(posterior=accepted.tail(10),
models=sampler_args['models'],
guess=sampler_args['guess'],
data=data_mu,
plot_index=plot_index,
titles=titles,
figs=[fig],
AX=AX)
ta.plot_steps({1: accepted}, skip=sampler_args['fixed_parameters'])
trace_figs = [plt.figure() for i in range(n_figs)]
trace_AX_ = [
trace_figs[i].add_subplot(5, 2, ii + 1) for i in range(len(trace_figs))
for ii in range(10)
]
trace_AX = dict(zip(trace_params, trace_AX_))
print('\ntrace_params:\n', trace_params)
print('\ntrace_AX:\n', trace_AX)
print('\ntraces:\n', traces)
#export the traces for all parameters into csv for ease of visualization
traces[1].to_csv(output_folder / 'Traces_dataframe.csv')
trace_figs, trace_AX = ta.plot_steps(traces,
skip=sampler_args['fixed_parameters'],
legend_args=legend_args,
figs=trace_figs,
AX=trace_AX)
print('\ntrace_AX\n', trace_AX)
#modify the figure setting else all the labels overlapped.
for tax in trace_AX.values():
tax.set_xticks(tax.get_xticks()[::2])
tax.set_yticks(tax.get_yticks()[::2])
tax.tick_params(axis='both', labelsize=16)
tax.xaxis.label.set_size(fontsize=16)
tax.yaxis.label.set_size(fontsize=16)
plt.tight_layout(pad=1.0) # set spacing between figures
#Rank models
table = ac.calculate_aic(data=sampler_args['data'],
'Fluor/OD': 'Pep',
'm': 'mRNA',
'indi': 'Inducer(Internal)'
}
}
labels = {
1: inducer_conc_str,
2: inducer_conc_str,
3: inducer_conc_str,
4: inducer_conc_str
}
legend_args = {'loc': 'upper left'}
figs, AX = cf.plot(posterior=posterior,
models=sampler_args['models'],
data=data_mu,
data_sd=data_sd,
plot_index=plot_index,
labels=labels,
titles=titles,
legend_args=legend_args,
figs=None,
AX=None)
for model_num in sampler_args['models']:
params = sampler_args['models'][model_num]['params']
traces_ = {key: traces[key][params] for key in traces}
figs, AX = ta.plot_steps(traces_,
skip=sampler_args['fixed_parameters'],
legend_args=legend_args)
seeder = seed(sampler_args['guess'], sampler_args['fixed_parameters'],
sampler_args['bounds'])
for i in range(10):
print('Run ' + str(i + 1))
sampler_args['guess'] = seeder()
result = cf.simulated_annealing(**sampler_args)
accepted = result['a']
traces[i + 1] = accepted.iloc[::5] #Thin the trace
for model_num in sampler_args['models']:
params = sampler_args['models'][model_num]['params']
traces_ = {key: traces[key][params] for key in traces}
t_figs, t_AX = ta.plot_steps(traces_,
skip=sampler_args['fixed_parameters'],
legend_args=legend_args)
k_figs, k_AX = ta.plot_steps(traces_,
skip=sampler_args['fixed_parameters'],
legend_args=legend_args)
h_figs, h_AX = ta.plot_steps(traces_,
skip=sampler_args['fixed_parameters'],
legend_args=legend_args)
'''
To save time, we recommend you thin the trace at your discretion!
'''
'''
9. A Priori Identifiability Analysis
'''
'''