def get_non_outliers(v, MAD_threshold=2.5):
""" returns the the non-zero, non-outliers for the input pandas series using MAD"""
outliers = get_list_of_outliers(v, threshold=MAD_threshold)
# setting the zero positions as outliers
outliers[np.where(v==0)] = 1
non_outliers = np.logical_not(outliers)
# The non-outliers are non-zero values that are in the IQR (positions that are zero are considered outliers
# even if the IQR includes zero)
non_outliers_indices = np.where(non_outliers)[0]
mean = np.mean(v[non_outliers_indices])
std = np.std(v[non_outliers_indices])
return non_outliers_indices, mean, std
def get_non_outliers_information(v, MAD_threshold=2.5):
""" returns the non-outliers for the input pandas series using MAD"""
d = pd.Series(index=columns_for_samples_coverage_stats_dict)
outliers = get_list_of_outliers(v, threshold=MAD_threshold)
non_outliers = np.logical_not(outliers)
if not (len(non_outliers)):
non_outlier_indices = np.array([])
d['non_outlier_mean_coverage'] = 0.0
d['non_outlier_coverage_std'] = 0.0
else:
non_outlier_indices = np.where(non_outliers)[0]
d['non_outlier_mean_coverage'] = np.mean(v[non_outlier_indices])
d['non_outlier_coverage_std'] = np.std(v[non_outlier_indices])
return non_outlier_indices, d
def get_non_outliers_information(v, MAD_threshold=2.5, zeros_are_outliers=False):
""" returns the non-outliers for the input pandas series using MAD"""
d = pd.Series(index=columns_for_samples_coverage_stats_dict)
outliers = get_list_of_outliers(v, threshold=MAD_threshold, zeros_are_outliers=zeros_are_outliers)
non_outliers = np.logical_not(outliers)
non_outlier_indices = np.where(non_outliers)[0]
if not(len(non_outlier_indices)):
non_outlier_indices = np.array([])
d['non_outlier_mean_coverage'] = 0.0
d['non_outlier_coverage_std'] = 0.0
else:
d['non_outlier_mean_coverage'] = np.mean(v[non_outlier_indices])
d['non_outlier_coverage_std'] = np.std(v[non_outlier_indices])
return non_outlier_indices, d