Merge branch 'master' of gitlab.fbk.eu:MPBA/inf_revamped

.gitignore

@@ -7,6 +7,7 @@ __pycache__/
 
 # Jupyter Notebook
 .ipynb_checkpoints
+*ipynb
 
 # VS code settings
 .vscode

sklearn_rf_training_fixrank.py

@@ -8,25 +8,24 @@ import numpy as np
 import pandas as pd
 import csv
 import os.path
-from scaling import norm_l2
 from mlpy import borda_count, canberra_stability
 from input_output import load_data
 import performance as perf
 import sys
-import tarfile
 import glob
 import argparse
 import configparser as ConfigParser
 from sklearn.ensemble import RandomForestClassifier
+from sklearn.multiclass import OneVsRestClassifier
 from sklearn.metrics import roc_auc_score, matthews_corrcoef, accuracy_score
 from sklearn import preprocessing
 from sklearn.model_selection import StratifiedKFold, StratifiedShuffleSplit, train_test_split
 import bootstrapped.bootstrap as bs
 import bootstrapped.stats_functions as bs_stats
 
-__author__  = 'Marco Chierici, Alessandro Zandona'
-__version__ = '2.0'
-__date__    = '15 December 2016'
+__author__  = 'Marco Chierici'
+__version__ = '2.5'
+__date__    = '28 Nov 2019'
 
 class myArgumentParser(argparse.ArgumentParser):
     def __init__(self, *args, **kwargs):
@@ -51,7 +50,6 @@ parser.add_argument('--cv_k', type=np.int, default=5, help='Number of CV folds (
 parser.add_argument('--cv_n', type=np.int, default=10, help='Number of CV cycles (default: %(default)s)')
 parser.add_argument('--rankFeats', type=str, default='', help='Ranked features list to be used by Machine Learning [Feats name on 1st column, feats weights on 2nd column, with HEADER]')
 parser.add_argument('--reliefk', type=np.int, default=3, help='Number of nearest neighbors for ReliefF (default: %(default)s)')
-#parser.add_argument('--plot', action='store_true', help='Plot metric values over all training cycles' )
 
 if len(sys.argv)==1:
     parser.print_help()
@@ -62,7 +60,6 @@ DATAFILE = args.DATAFILE
 LABELSFILE = args.LABELSFILE
 RANK_METHOD = args.RANK_METHOD
 OUTDIR = args.OUTDIR
-#plot_out = args.plot
 random_labels = args.random
 CV_K = args.cv_k
 CV_N = args.cv_n
@@ -70,8 +67,8 @@ RANKFEATS = args.rankFeats
 relief_k = args.reliefk
 
 BASEFILE = os.path.splitext(os.path.basename(DATAFILE))[0]
-SVM_TYPE = 'RandomForest'
-OUTFILE = os.path.join(OUTDIR, '_'.join([BASEFILE, SVM_TYPE, 'RF' if RANK_METHOD=='randomForest' else RANK_METHOD]))
+MODEL_TYPE = 'RandomForest'
+OUTFILE = os.path.join(OUTDIR, '_'.join([BASEFILE, MODEL_TYPE, 'RF' if RANK_METHOD=='randomForest' else RANK_METHOD]))
 
 # create OUTDIR if not present
 try:
@@ -84,7 +81,7 @@ except OSError:
 if RANK_METHOD == 'ReliefF':
     from relief import ReliefF
     # add ReliefF K to OUTFILE
-    OUTFILE = os.path.join(OUTDIR, '_'.join([BASEFILE, SVM_TYPE, RANK_METHOD + str(relief_k)]))
+    OUTFILE = os.path.join(OUTDIR, '_'.join([BASEFILE, MODEL_TYPE, RANK_METHOD + str(relief_k)]))
 elif RANK_METHOD == 'tree' :
     from sklearn.ensemble import ExtraTreesClassifier
 elif RANK_METHOD == 'KBest':
@@ -96,8 +93,11 @@ TUN_CV_K = 10
 TUN_CV_P = 50
 
 sample_names, var_names, x = load_data(DATAFILE)
-y = np.loadtxt(LABELSFILE, dtype=np.int)
-#print(var_names[:10])
+y_orig = np.loadtxt(LABELSFILE, dtype=np.int)
+# encode labels
+le = preprocessing.LabelEncoder()
+y = le.fit_transform(y_orig)
+
 # If ranked list is given as input to DAP, read it and extract features index
 if RANK_METHOD == "rankList":
     rankedList = np.loadtxt(RANKFEATS, delimiter='\t', dtype=str, skiprows=1)
@@ -167,7 +167,7 @@ for n in range(CV_N):
         elif RANK_METHOD == 'randomForest' :
             ranking_tmp = np.argsort(forest.feature_importances_)[::-1]
         elif RANK_METHOD == 'KBest':
-            selector = SelectKBest(f_classif)
+            selector = SelectKBest(f_classif, k='all')
             selector.fit(x_tr, y_tr)
             ranking_tmp = np.argsort( -np.log10(selector.pvalues_) )[::-1]
         elif RANK_METHOD == 'rankList':
@@ -279,14 +279,3 @@ config.set("OUTPUT", "MCC", np.max(AMCC))
 config.set("OUTPUT", "N_feats", opt_feats)
 config.write(logf)
 logf.close()
-
-
-#if plot_out:
-#    from metrics_plot import *
-#    plt_title = (' ').join( [os.path.basename(DATAFILE).replace('.txt', ''), SVM_TYPE] )
-#    if random_labels:
-#        metplot(RLFile = (OUTFILE + "_metrics.txt"), title = plt_title)
-#    elif RANK_METHOD=='random':
-#        metplot(RRFile = (OUTFILE + "_metrics.txt"), title = plt_title)
-#    else: 
-#        metplot(normFile = (OUTFILE + "_metrics.txt"), title = plt_title)

sklearn_rf_validation_writeperf.py

@@ -11,6 +11,8 @@ import sys
 
 import numpy as np
 from sklearn.ensemble import RandomForestClassifier
+from sklearn.metrics import roc_auc_score, matthews_corrcoef, accuracy_score
+from sklearn.preprocessing import LabelEncoder
 
 import performance as perf
 from extract_topfeats import extract_feats
@@ -59,13 +61,18 @@ extract_feats(TRFILE, RANK, NFEATS, TR_TOPFEATS)
 TS_TOPFEATS = OUTFILE + '_top%s_ts.txt' % NFEATS
 extract_feats(TSFILE, RANK, NFEATS, TS_TOPFEATS)
 
+# initialize LabelEncoder
+le = LabelEncoder()
+
 # load data
 sample_names_tr, var_names_tr, x_tr = load_data(TR_TOPFEATS)
 y_tr = np.loadtxt(LABELSFILE, dtype=np.int, delimiter='\t')
+y_tr = le.fit_transform(y_tr)
 sample_names_ts, var_names_ts, x_ts = load_data(TS_TOPFEATS)
 # load the TS labels if available
 if TSLABELSFILE is not None:
     y_ts = np.loadtxt(TSLABELSFILE, dtype=np.int, delimiter='\t')
+    y_ts = le.transform(y_ts)
 
 # prediction
 forest = RandomForestClassifier(n_estimators=500, criterion='gini', random_state=0)
@@ -74,26 +81,30 @@ forest.fit(x_tr, y_tr)
 p_tr = forest.predict(x_tr)
 p_ts = forest.predict(x_ts)
 
+# decode labels back
+p_tr_dec = le.inverse_transform(p_tr)
+p_ts_dec = le.inverse_transform(p_ts)
+
 prob_tr = forest.predict_proba(x_tr)
 prob_ts = forest.predict_proba(x_ts)
 
-print("MCC on train: %.3f" % (perf.KCCC_discrete(y_tr, p_tr)))
+print("MCC on train: %.3f" % (matthews_corrcoef(y_tr, p_tr)))
 if TSLABELSFILE is not None:
-    print("MCC on validation: %.3f" % (perf.KCCC_discrete(y_ts, p_ts)))
+    print("MCC on validation: %.3f" % (matthews_corrcoef(y_ts, p_ts)))
 
 # write output files
 # save MCC_train and MCC_validation
 with open(OUTFILE + "_MCC_scores.txt", "w") as fout:
-    fout.write("MCC_train\t%.5f\n" % (perf.KCCC_discrete(y_tr, p_tr)))
-    fout.write("MCC_validation\t%.5f\n" % (perf.KCCC_discrete(y_ts, p_ts)))
+    fout.write("MCC_train\t%.5f\n" % (matthews_corrcoef(y_tr, p_tr)))
+    fout.write("MCC_validation\t%.5f\n" % (matthews_corrcoef(y_ts, p_ts)))
 
 with open(OUTFILE + "_TEST_pred_tr.txt", "w") as fout:
     for i in range(len(sample_names_tr)):
-        fout.write("%s\t%i\n" % (sample_names_tr[i], p_tr[i]))
+        fout.write("%s\t%i\n" % (sample_names_tr[i], p_tr_dec[i]))
 
 with open(OUTFILE + "_TEST_pred_ts.txt", "w") as fout:
     for i in range(len(sample_names_ts)):
-        fout.write("%s\t%i\n" % (sample_names_ts[i], p_ts[i]))
+        fout.write("%s\t%i\n" % (sample_names_ts[i], p_ts_dec[i]))
 
 np.savetxt(OUTFILE + "_TEST_signature.txt",
            np.array(var_names_tr).reshape(-1,1),