def stats_overall(self):
"""
Returns an OrderedDict with overall statistics
"""
df = self._df_confusion
d_stats = collections.OrderedDict()
d_class_agreement = class_agreement(df)
key = 'Accuracy'
try:
d_stats[key] = d_class_agreement['diag'] # 0.35
except KeyError:
d_stats[key] = np.nan
key = '95% CI'
try:
d_stats[key] = binom_interval(np.sum(np.diag(df)),
df.sum().sum()) # (0.1539, 0.5922)
except: # noqa
d_stats[key] = np.nan
d_prop_test = prop_test(df)
d_stats['No Information Rate'] = 'ToDo' # 0.8
d_stats['P-Value [Acc > NIR]'] = d_prop_test['p.value'] # 1
d_stats['Kappa'] = d_class_agreement['kappa'] # 0.078
d_stats['Mcnemar\'s Test P-Value'] = 'ToDo' # np.nan
return (d_stats)
def stats_overall(self):
"""
Returns an OrderedDict with overall statistics
"""
df = self._df_confusion
d_stats = collections.OrderedDict()
d_class_agreement = class_agreement(df)
key = 'Accuracy'
try:
d_stats[key] = d_class_agreement['diag'] # 0.35
except:
d_stats[key] = np.nan
key = '95% CI'
try:
d_stats[key] = binom_interval(np.sum(np.diag(df)), df.sum().sum()) # (0.1539, 0.5922)
except:
d_stats[key] = np.nan
d_prop_test = prop_test(df)
d_stats['No Information Rate'] = 'ToDo' # 0.8
d_stats['P-Value [Acc > NIR]'] = d_prop_test['p.value'] # 1
d_stats['Kappa'] = d_class_agreement['kappa'] # 0.078
d_stats['Mcnemar\'s Test P-Value'] = 'ToDo' # np.nan
return(d_stats)