def _get_stats_for_state(deaths, pop):
current_week = deaths[-7:]
last_week = deaths[-14:-7]
last_week_2 = deaths[-21:-14]
the_dict = {}
resample1, resample2 = bootstrap.resample_two_samples(last_week, current_week, num_iterations = 1000)
both = bootstrap.combine_resamples(last_week, current_week, resample1, resample2)
p_value = bootstrap.get_p_value(both)
resample3, resample4 = bootstrap.resample_two_samples(last_week_2, last_week, num_iterations = 1000)
both2 = bootstrap.combine_resamples(last_week_2, current_week, resample3, resample4)
p_value2 = bootstrap.get_p_value(both2)
the_dict['p_value_last_week'] = p_value
the_dict['p_value_last_week2'] = p_value2
the_dict['current_week_mean'] = np.mean(current_week)
the_dict['last_week_mean'] = np.mean(last_week)
the_dict['last_week_mean_2'] = np.mean(last_week_2)
if pop == None:
the_dict['current_week_per_million'] = None
the_dict['last_week_per_million'] = None
else:
the_dict['current_week_per_million'] = round(float(np.mean(current_week))/pop * 1000000, 1)
the_dict['last_week_per_million'] = round(float(np.mean(last_week))/pop * 1000000, 1)
return the_dict
def _get_stats_for_state(deaths, cases, pop):
current_week = deaths[-7:]
last_week = deaths[-14:-7]
last_week_2 = deaths[-21:-14]
current_week_cases = cases[-7:]
last_week_cases = cases[-14:-7]
last_week_2_cases = cases[-21:-14]
the_dict = {}
for i in [
(last_week, current_week, 'p_value_death_current_last'),
(last_week_2, current_week, 'p_value_death_last_last2'),
(last_week_cases, current_week_cases, 'p_value_cases_current_last'),
(last_week_2_cases, current_week_cases, 'p_value_cases_last_last2'),
]:
resample1, resample2 = bootstrap.resample_two_samples(
i[0], i[1], num_iterations=1000)
both = bootstrap.combine_resamples(last_week, current_week, resample1,
resample2)
p_value = bootstrap.get_p_value(both)
the_dict[i[2]] = p_value
the_dict['current_week_mean'] = np.mean(current_week)
the_dict['last_week_mean'] = np.mean(last_week)
the_dict['last_week_mean_2'] = np.mean(last_week_2)
the_dict['current_week_cases_mean'] = np.mean(current_week_cases)
the_dict['last_week_cases_mean'] = np.mean(last_week_cases)
the_dict['last_week_cases_mean_2'] = np.mean(last_week_2_cases)
if pop == None:
the_dict['current_week_per_million'] = None
the_dict['last_week_per_million'] = None
the_dict['current_week_cases_per_million'] = None
the_dict['last_week_cases_per_million'] = None
else:
the_dict['current_week_per_million'] = round(
float(np.mean(current_week)) / pop * 1000000, 1)
the_dict['last_week_per_million'] = round(
float(np.mean(last_week)) / pop * 1000000, 1)
the_dict['current_week_cases_per_million'] = round(
float(np.mean(current_week_cases)) / pop * 1000000, 1)
the_dict['last_week_cases_per_million'] = round(
float(np.mean(last_week_cases)) / pop * 1000000, 1)
return the_dict
def test_pvalue(self):
sample1 = [1, 2, 3, 5]
sample2 = [1, 2, 3, 5]
r1, r2 = bootstrap.resample_two_samples(sample1, sample2)
b = bootstrap.combine_resamples(sample1, sample2, r1, r2)
bootstrap.get_p_value(b)