Fix some cv prediction stuff

tests/test_cv_full.py

@@ -10,6 +10,7 @@ import util_functions
 import essay_set
 import feature_extractor
 import numpy
+import math
 
 from sklearn.ensemble import GradientBoostingClassifier
 
@@ -40,7 +41,7 @@ for filename in filenames:
     extractor.initialize_dictionaries(eset)
     train_feats=extractor.gen_feats(eset)
     clf=GradientBoostingClassifier(n_estimators=100, learn_rate=.05,max_depth=4, random_state=1,min_samples_leaf=3)
-    cv_preds=util_functions.gen_cv_preds(clf,train_feats,scores)
+    cv_preds=util_functions.gen_cv_preds(clf,train_feats,scores, num_chunks = int(math.floor(len(texts)/2)))
     err=numpy.mean(numpy.abs(cv_preds-scores))
     print err
     kappa=util_functions.quadratic_weighted_kappa(list(cv_preds),scores)

util_functions.py

@@ -269,12 +269,12 @@ def gen_cv_preds(clf, arr, sel_score, num_chunks=3):
     preds = []
     set_score = numpy.asarray(sel_score, dtype=numpy.int)
     chunk_vec = numpy.asarray(range(0, len(chunks)))
-    for i in range(0, len(chunks)):
+    for i in xrange(0, len(chunks)):
         loop_inds = list(
             chain.from_iterable([chunks[int(z)] for z, m in enumerate(range(0, len(chunks))) if int(z) != i]))
         sim_fit = clf.fit(arr[loop_inds], set_score[loop_inds])
-        preds.append(sim_fit.predict(arr[chunks[i]]))
-    all_preds = numpy.concatenate((preds[0], preds[1], preds[2]), axis=0)
+        preds.append(list(sim_fit.predict(arr[chunks[i]])))
+    all_preds = list(chain(*preds))
     return(all_preds)