def _statistic_summary(table,
input_cols,
statistics,
column_indices=None,
percentile_amounts=None,
trimmed_mean_amounts=None):
tmp_table = table.copy()
tmp_table = list(map(list, zip(*tmp_table)))
result = []
for i in range(len(input_cols)):
data = [input_cols[i]]
for st in statistics:
if 'max' == st:
data.append(brtc_stats.max(tmp_table[column_indices[i]]))
elif 'min' == st:
data.append(brtc_stats.min(tmp_table[column_indices[i]]))
elif 'range' == st:
data.append(brtc_stats.range(tmp_table[column_indices[i]]))
elif 'sum' == st:
data.append(brtc_stats.sum(tmp_table[column_indices[i]]))
elif 'avg' == st:
data.append(brtc_stats.mean(tmp_table[column_indices[i]]))
elif 'variance' == st:
data.append(brtc_stats.var_samp(tmp_table[column_indices[i]]))
elif 'stddev' == st:
data.append(brtc_stats.std(tmp_table[column_indices[i]]))
elif 'skewness' == st:
data.append(brtc_stats.skewness(tmp_table[column_indices[i]]))
elif 'kurtosis' == st:
data.append(brtc_stats.kurtosis(tmp_table[column_indices[i]]))
elif 'nrow' == st:
data.append(brtc_stats.num_row(tmp_table[column_indices[i]]))
elif 'num_of_value' == st:
data.append(brtc_stats.num_value(tmp_table[column_indices[i]]))
elif 'null_count' == st:
data.append(brtc_stats.num_null(tmp_table[column_indices[i]]))
elif 'mode' == st:
data.append(brtc_stats.mode(tmp_table[column_indices[i]]))
elif 'median' == st:
data.append(brtc_stats.median(tmp_table[column_indices[i]]))
elif 'q1' == st:
data.append(brtc_stats.q1(tmp_table[column_indices[i]]))
elif 'q3' == st:
data.append(brtc_stats.q3(tmp_table[column_indices[i]]))
elif 'iqr' == st:
data.append(brtc_stats.iqr(tmp_table[column_indices[i]]))
elif 'percentile' == st and percentile_amounts is not None:
for pa in _unique_list(percentile_amounts):
data.append(
brtc_stats.percentile(tmp_table[column_indices[i]],
pa))
elif 'trimmed_mean' == st and trimmed_mean_amounts is not None:
for ta in _unique_list(trimmed_mean_amounts):
data.append(
brtc_stats.trimmed_mean(tmp_table[column_indices[i]],
ta))
result.append(data)
return {'out_table': result}
def _statistic_summary_original(table, input_cols, statistics, percentile_amounts=None, trimmed_mean_amounts=None):
_table = table.copy()
data = {'column_name':input_cols}
for st in statistics:
if 'max' == st:
data['max'] = [brtc_stats.max(_table[x]) for x in input_cols]
if 'min' == st:
data['min'] = [brtc_stats.min(_table[x]) for x in input_cols]
if 'range' == st:
data['range'] = [brtc_stats.range(_table[x]) for x in input_cols]
if 'sum' == st:
data['sum'] = [brtc_stats.sum(_table[x]) for x in input_cols]
if 'avg' == st:
data['avg'] = [brtc_stats.mean(_table[x]) for x in input_cols]
if 'variance' == st:
data['variance'] = [brtc_stats.var_samp(_table[x]) for x in input_cols]
if 'stddev' == st:
data['stddev'] = [brtc_stats.std(_table[x]) for x in input_cols]
if 'skewness' == st:
data['skewness'] = [brtc_stats.skewness(_table[x]) for x in input_cols]
if 'kurtosis' == st:
data['kurtosis'] = [brtc_stats.kurtosis(_table[x]) for x in input_cols]
if 'nrow' == st:
data['num_of_row'] = [brtc_stats.num_row(_table[x]) for x in input_cols]
if 'num_of_value' == st:
data['num_of_value'] = [brtc_stats.num_value(_table[x]) for x in input_cols]
if 'null_count' == st:
data['null_count'] = [brtc_stats.num_null(_table[x]) for x in input_cols]
if 'mode' == st:
data['mode'] = [list(brtc_stats.mode(_table[x])) for x in input_cols]
if 'median' == st:
data['median'] = [brtc_stats.median(_table[x]) for x in input_cols]
if 'q1' == st:
data['q1'] = [brtc_stats.q1(_table[x]) for x in input_cols]
if 'q3' == st:
data['q3'] = [brtc_stats.q3(_table[x]) for x in input_cols]
if 'iqr' == st:
data['iqr'] = [brtc_stats.iqr(_table[x]) for x in input_cols]
if 'percentile' == st and percentile_amounts is not None:
for pa in _unique_list(percentile_amounts):
pa_colname = 'percentile_{}'.format(_amounts_colname(pa))
data[pa_colname] = [brtc_stats.percentile(_table[x], pa) for x in input_cols]
if 'trimmed_mean' == st and trimmed_mean_amounts is not None:
for ta in _unique_list(trimmed_mean_amounts):
ta_colname = 'trimmed_mean_{}'.format(_amounts_colname(ta))
data[ta_colname] = [brtc_stats.trimmed_mean(_table[x], ta) for x in input_cols]
result = pd.DataFrame(data)
return {'out_table' : result}
def _statistic_summary(table,
input_cols,
statistics,
percentile_amounts=[],
trimmed_mean_amounts=[]):
_table = table.copy()
percentile_amounts = np.unique(percentile_amounts)
trimmed_mean_amounts = np.unique(trimmed_mean_amounts)
def _amounts_colname(a):
return str(a).replace('.', '_')
data = {'column_name': input_cols}
for st in statistics:
if 'max' == st:
data['max'] = [brtc_stats.max(_table[x]) for x in input_cols]
if 'min' == st:
data['min'] = [brtc_stats.min(_table[x]) for x in input_cols]
if 'range' == st:
data['range'] = [brtc_stats.range(_table[x]) for x in input_cols]
if 'sum' == st:
data['sum'] = [brtc_stats.sum(_table[x]) for x in input_cols]
if 'avg' == st:
data['avg'] = [brtc_stats.mean(_table[x]) for x in input_cols]
if 'variance' == st:
data['variance'] = [
brtc_stats.var_samp(_table[x]) for x in input_cols
]
if 'stddev' == st:
data['stddev'] = [brtc_stats.std(_table[x]) for x in input_cols]
if 'skewness' == st:
data['skewness'] = [
brtc_stats.skewness(_table[x]) for x in input_cols
]
if 'kurtosis' == st:
data['kurtosis'] = [
brtc_stats.kurtosis(_table[x]) for x in input_cols
]
if 'nrow' == st:
data['nrow'] = [brtc_stats.num_row(_table[x]) for x in input_cols]
if 'num_of_value' == st:
data['num_of_value'] = [
brtc_stats.num_value(_table[x]) for x in input_cols
]
# if 'nan_count' == st:
# data['nan_count'] = [brtc_stats.num_nan(_table[x]) for x in input_cols]
if 'null_count' == st:
data['null_count'] = [
brtc_stats.num_null(_table[x]) for x in input_cols
]
if 'mode' == st:
data['mode'] = [brtc_stats.mode(_table[x]) for x in input_cols]
if 'median' == st:
data['median'] = [brtc_stats.median(_table[x]) for x in input_cols]
if 'q1' == st:
data['q1'] = [brtc_stats.q1(_table[x]) for x in input_cols]
if 'q3' == st:
data['q3'] = [brtc_stats.q3(_table[x]) for x in input_cols]
if 'iqr' == st:
data['iqr'] = [brtc_stats.iqr(_table[x]) for x in input_cols]
if 'percentile' == st:
for pa in percentile_amounts:
pa_colname = 'percentile_{}'.format(_amounts_colname(pa))
data[pa_colname] = [
brtc_stats.percentile(_table[x], pa) for x in input_cols
]
if 'trimmed_mean' == st:
for ta in trimmed_mean_amounts:
ta_colname = 'trimmed_mean_{}'.format(_amounts_colname(ta))
data[ta_colname] = [
brtc_stats.trimmed_mean(_table[x], ta) for x in input_cols
]
result = pd.DataFrame(data)
return {'out_table': result}