def cdf_plot(model, timeline=None, ax=None, **plot_kwargs):
"""
"""
from lifelines import KaplanMeierFitter
if ax is None:
ax = plt.gca()
if timeline is None:
timeline = model.timeline
COL_EMP = "empirical CDF"
if CensoringType.is_left_censoring(model):
empirical_kmf = KaplanMeierFitter().fit_left_censoring(
model.durations, model.event_observed, label=COL_EMP, timeline=timeline
)
elif CensoringType.is_right_censoring(model):
empirical_kmf = KaplanMeierFitter().fit_right_censoring(
model.durations, model.event_observed, label=COL_EMP, timeline=timeline
)
elif CensoringType.is_interval_censoring(model):
raise NotImplementedError("lifelines does not have a non-parametric interval model yet.")
empirical_kmf.plot_cumulative_density(ax=ax, **plot_kwargs)
dist = get_distribution_name_of_lifelines_model(model)
dist_object = create_scipy_stats_model_from_lifelines_model(model)
ax.plot(timeline, dist_object.cdf(timeline), label="fitted %s" % dist, **plot_kwargs)
ax.legend()
return ax
def cdf_plot(model, timeline=None, **plot_kwargs):
from lifelines import KaplanMeierFitter
set_kwargs_ax(plot_kwargs)
ax = plot_kwargs.pop("ax")
if timeline is None:
timeline = model.timeline
kmf = KaplanMeierFitter().fit(
model.durations,
model.event_observed,
left_censorship=model.left_censorship,
label="empirical CDF",
timeline=timeline,
)
if model.left_censorship:
kmf.plot_cumulative_density(ax=ax, **plot_kwargs)
else:
(1 - kmf.survival_function_).plot(ax=ax, **plot_kwargs)
dist = get_distribution_name_of_lifelines_model(model)
dist_object = create_scipy_stats_model_from_lifelines_model(model)
ax.plot(timeline,
dist_object.cdf(timeline),
label="fitted %s" % dist,
**plot_kwargs)
ax.legend()
return ax
def cdf_plot(model, timeline=None, **plot_kwargs):
from lifelines import KaplanMeierFitter
set_kwargs_ax(plot_kwargs)
ax = plot_kwargs.pop("ax")
if timeline is None:
timeline = model.timeline
COL_EMP = "empirical quantiles"
if CensoringType.is_left_censoring(model):
kmf = KaplanMeierFitter().fit_left_censoring(model.durations,
model.event_observed,
label=COL_EMP,
timeline=timeline)
elif CensoringType.is_right_censoring(model):
kmf = KaplanMeierFitter().fit_right_censoring(model.durations,
model.event_observed,
label=COL_EMP,
timeline=timeline)
elif CensoringType.is_interval_censoring(model):
raise NotImplementedError()
kmf.plot_cumulative_density(ax=ax, **plot_kwargs)
dist = get_distribution_name_of_lifelines_model(model)
dist_object = create_scipy_stats_model_from_lifelines_model(model)
ax.plot(timeline,
dist_object.cdf(timeline),
label="fitted %s" % dist,
**plot_kwargs)
ax.legend()
return ax
def cdf_plot(model, timeline=None, ax=None, **plot_kwargs):
"""
This plot compares the empirical CDF (derived by KaplanMeier) vs the model CDF.
Parameters
------------
model: lifelines univariate model
timeline: iterable
ax: matplotlib axis
"""
from lifelines import KaplanMeierFitter
from matplotlib import pyplot as plt
if ax is None:
ax = plt.gca()
if timeline is None:
timeline = model.timeline
COL_EMP = "empirical CDF"
if CensoringType.is_left_censoring(model):
empirical_kmf = KaplanMeierFitter().fit_left_censoring(
model.durations,
model.event_observed,
label=COL_EMP,
timeline=timeline,
weights=model.weights,
entry=model.entry)
elif CensoringType.is_right_censoring(model):
empirical_kmf = KaplanMeierFitter().fit_right_censoring(
model.durations,
model.event_observed,
label=COL_EMP,
timeline=timeline,
weights=model.weights,
entry=model.entry)
elif CensoringType.is_interval_censoring(model):
empirical_kmf = KaplanMeierFitter().fit_interval_censoring(
model.lower_bound,
model.upper_bound,
label=COL_EMP,
timeline=timeline,
weights=model.weights,
entry=model.entry)
empirical_kmf.plot_cumulative_density(ax=ax, **plot_kwargs)
dist = get_distribution_name_of_lifelines_model(model)
dist_object = create_scipy_stats_model_from_lifelines_model(model)
ax.plot(timeline,
dist_object.cdf(timeline),
label="fitted %s" % dist,
**plot_kwargs)
ax.legend()
return ax
ax1.step(kmf.cumulative_density_.index.values,
kmf.cumulative_density_.values,
color=lab_colors[i],
lw=2)
kmf.fit(training_time['sessions'].values,
event_observed=training_time['trained'])
ax1.step(kmf.cumulative_density_.index.values,
kmf.cumulative_density_.values,
color='black',
lw=2)
ax1.set(ylabel='Cumulative probability of\nreaching trained criterion',
xlabel='Training day',
xlim=[0, 60],
ylim=[0, 1.02])
kmf.fit(training_time['sessions'].values,
event_observed=training_time['trained'])
kmf.plot_cumulative_density(ax=ax2)
ax2.set(ylabel='Cumulative probability of\nreaching trained criterion',
xlabel='Training day',
title='All labs',
xlim=[0, 60],
ylim=[0, 1.02])
ax2.get_legend().set_visible(False)
sns.despine(trim=True, offset=5)
plt.tight_layout(pad=2)
seaborn_style()
plt.savefig(join(fig_path, 'figure2c_probability_trained.pdf'), dpi=300)
plt.savefig(join(fig_path, 'figure2c_probability_trained.png'), dpi=300)
#Plot survival function with confidence interval:
confidence_surv_func = kmf.confidence_interval_survival_function_
plt.plot(confidence_surv_func["KM_estimate_lower_0.95"],label="Lower")
plt.plot(confidence_surv_func["KM_estimate_upper_0.95"],label="Upper")
plt.title("Survival Function With Confidence Interval")
plt.xlabel("Number of days")
plt.ylabel("Survival Probability")
plt.legend()
#Probabaility of a subject dying:
#p(1022) = p(0) +......+p(1022)
kmf.cumulative_density_
#Plot the cumulative density graph:
kmf.plot_cumulative_density()
plt.title("Cumulative Density Plot")
plt.xlabel("Number of days")
plt.ylabel("Probability of person's death")
#Cumulative density with confidence interval:
kmf.confidence_interval_cumulative_density_
#Plot cumulative density with confidence interval:
confidence_cumulative_density = kmf.confidence_interval_cumulative_density_
plt.plot(kmf.confidence_interval_cumulative_density_["KM_estimate_lower_0.95"],label="Lower")
plt.plot(kmf.confidence_interval_cumulative_density_["KM_estimate_upper_0.95"],label="Upper")
plt.title("Cumulative Density With Confidence Interval")
plt.xlabel("Number of days")
plt.ylabel("Cumulative Density")
plt.legend()
#Plot the survival_function data:
kmf_m.plot()
kmf_f.plot()
plt.xlabel("Days Passed")
plt.ylabel("Survival Probability")
plt.title("KMF")
#Cumulative density for male group:
kmf_m.cumulative_density_
#Cumulative density for female group:
kmf_f.cumulative_density_
#PLot the graph for cumulative density for both groups:
kmf_m.plot_cumulative_density()
kmf_f.plot_cumulative_density()
plt.title("Cumulative Density")
plt.xlabel("Number of days")
plt.ylabel("Probability")
#Hazard Function:
from lifelines import NelsonAalenFitter
#Fitting the data into objects:
naf_m = NelsonAalenFitter()
naf_f = NelsonAalenFitter()
naf_m.fit(Male["time"], event_observed=Male["dead"])
naf_f.fit(Female["time"], event_observed=Female["dead"])
#Cumulative hazard for male group: