forked from APPIAN-PET/APPIAN
-
Notifications
You must be signed in to change notification settings - Fork 0
/
QC.py
152 lines (118 loc) · 6.83 KB
/
QC.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
import os
import pandas as pd
import numpy as np
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
import seaborn as sns
from glob import glob
from sys import argv
from re import sub,split
from src.qc import outlier_measures
from sklearn.metrics import roc_auc_score
from argparse import ArgumentParser
def read_df(out_dir, error_type='rot', df_fn='metric_df.csv', clobber=False) :
if not os.path.exists(df_fn) or clobber :
file_strings=out_dir + '/preproc/_*/*'+error_type+'*_qc_metrics*/*csv'
files = glob(file_strings)
df_list=[]
for fn in files :
df0 = pd.read_csv(fn)
for metric, df in df0.groupby(['metric']) :
fn_split= split('_|/', fn)
error_level = np.array( [ sub(error_type+'-', '', x).split('-') for x in fn_split if error_type+'-' in x ]).astype(float)[0]
for i, dim in enumerate(['x','y','z']) :
df2 = df.copy()
df2['error_type']=[error_type]*df2.shape[0]
df2['error_level']=error_level[i]
df2['axis']=dim
df_list.append(df2)
df = pd.concat(df_list)
df.to_csv(df_fn, index=False)
else :
df = pd.read_csv(df_fn)
return df
def outlier_factor(df, out_fn='outlier_df.csv', n_steps=10, axis_list=['x'], clobber=False):
if not os.path.exists(out_fn) or clobber :
df_list=[]
for (metric, axis), df_metric in df.groupby(['metric','axis']):
if not axis in axis_list : continue
df_no_error = df_metric.loc[ df_metric['error_level'] == 0 ]
df_error = df_metric.loc[ df_metric['error_level'] != -1 ]
for columns, df2 in df_error.groupby(['error_type','error_level','analysis','roi']):
for sub_idx, df_sub in df2.groupby(['sub_idx']) :
df_test=df_no_error.copy()
metric_list = df_test['value'].values
metric_list[ df_test['sub_idx'] == sub_idx ] = df_sub['value'].values
idx=np.arange(df_test.shape[0])[ df_test['sub_idx'] == sub_idx ].astype(int) #df_test[(df_test['sub'] == sub) & (df_test['ses']==ses) ].index[0]
for name, function in outlier_measures.items() :
df_sub['outlier_metric'] = name
outlier_values = function(metric_list)
if type( outlier_values[idx]) == np.ndarray or type( outlier_values[idx]) == list :
x = float(outlier_values[idx][0])
else :
x = outlier_values[idx]
#if name == 'KDE' :
# print(outlier_values.shape)
# print(type(outlier_values), type( outlier_values[idx]) == np.ndarray)
# print(outlier_values[idx])
# print(x)
df_sub['outlier_value'] = x
df_list.append(pd.DataFrame({'error_type':columns[0],'error_level':columns[1],'analysis':columns[2],'roi':columns[3],'sub_idx':[sub_idx], 'metric':[metric], 'metric_value':df_sub['value'], 'outlier':[name], 'outlier_value':[x], 'axis':[axis] }))
outlier_df = pd.concat(df_list)
outlier_df.to_csv(out_fn,index=False)
else :
outlier_df = pd.read_csv(out_fn)
return outlier_df
def auc(df,auc_fn='auc_df.csv', clobber=False):
if not os.path.exists(auc_fn) or clobber :
df_list=[]
for metric, df2 in df.groupby(['metric','outlier']):
df_no_error = df2.loc[ df2['error_level'] == 0 ]
df_error = df2.loc[ df2['error_level'] != -1 ]
for columns, df3 in df_error.groupby(['error_type','error_level','analysis','roi', 'metric', 'axis']):
for sub_ses, df_sub in df3.groupby(['sub_idx']) :
df_test=df_no_error.copy()
df_test.index = [2]*df_test.shape[0]
df_sub.index = [2]*df_sub.shape[0]
sub_idx = df_test['sub_idx'] == sub_ses
df_test.loc[sub_idx, : ] = df_sub
df_test.reset_index(inplace=True)
idx = df_test[df_test['sub_idx'] == sub_idx ].index[0]
y_predicted = df_test['outlier_value'].values
y_true = np.zeros(df_test.shape[0])
y_true[idx] = 1
auc = roc_auc_score(y_true, y_predicted)
df_list.append(pd.DataFrame({'error_type':columns[0],'error_level':columns[1],'analysis':columns[2],'roi':columns[3],'axis':[columns[5]],'sub_idx':[sub_idx], 'metric':df_sub['metric'], 'metric_value':df_sub['metric_value'], 'outlier':df_sub['outlier'], 'outlier_value':df_sub['outlier_value'],'auc':[auc] }))
auc_df = pd.concat(df_list)
auc_df.to_csv(auc_fn, index=False)
else :
auc_df = pd.read_csv(auc_fn)
return auc_df
def plot(df, out_fn='df.png', variables=['outlier','metric'], y='outlier_value', row='outlier', col='metric', hue='sub' ):
df_mean = df.groupby(['axis', 'error_type','analysis','roi'] + variables + ['error_level']).mean()
print(df_mean)
plt.figure(figsize=(8,12))
g = sns.FacetGrid(df, row=row, col=col, hue=hue, sharey=False)
g.map(plt.plot, "error_level", y)
plt.savefig(out_fn)
if __name__ == '__main__':
parser = ArgumentParser(usage="useage: ")
parser.add_argument("-s","--source","--sourcedir",dest="sourceDir", help="Path for input file directory", required=True)
parser.add_argument("-c","--clobber", dest="clobber", action='store_true', help="Clobber", default=False)
parser.add_argument("-a","--axis", dest="axis", type=str, help="Error Axis", default='x')
args = parser.parse_args()
out_dir = args.sourceDir
clobber= args.clobber
axis = args.axis
metric_df = read_df(out_dir, error_type='rot', clobber=clobber)
metric_df = metric_df.loc[metric_df['axis']==axis]
metric_df = metric_df.loc[metric_df['metric']=='MattesMutualInformation']
metric_df['sub_idx']= metric_df['sub'].astype(str) +'_'+metric_df['ses']
plot(metric_df, variables=['sub_idx'], out_fn='qc_metric.png', y='value', col='metric', row='analysis',hue='sub_idx' )
outlier_df = outlier_factor(metric_df, clobber=clobber)
print(outlier_df)
exit(0)
plot(outlier_df, variables=['sub_idx','metric'], out_fn='qc_outlier.png', y='outlier_value', col='outlier',row='metric', hue='sub_idx' )
auc_df = auc(outlier_df, clobber=clobber)
plot(auc_df, variables=['metric','outlier'], y='auc', col='outlier', hue='metric', out_fn='qc_auc.png' )