diff --git a/speech_data.py b/speech_data.py
index 25ea1e2..2d1bb2c 100644
--- a/speech_data.py
+++ b/speech_data.py
@@ -54,15 +54,32 @@ def sunflower_pairs_data():
     te_y = np.array(x_pos_test.shape[0]*[[1,0]])
     tr_pairs = np.array([x_pos_train,x_neg_train]).reshape(x_pos_train.shape[0],2,max_samples,sample_size)
     te_pairs = np.array([x_pos_test,x_neg_test]).reshape(x_pos_test.shape[0],2,max_samples,sample_size)
-    # x_data.shape
-    # y_data.shape
-    # train_test_split(x_data,y_data,test_size=0.33)[].shape
-    # len(train_test_split(x_data,y_data,test_size=0.33))
-    # sunflowers.loc[:,'file'][0]
-    # generate_aiff_spectrogram('outputs/sunflowers-Alex-150-normal-589.aiff')
-    # sunflowers[sunflowers['variant'] == 'phoneme']
-    # sunflowers[sunflowers['variant'] == 'normal']
-    # for s in sunflowers.values:
-    #     print(s)
-    #return train_test_split(x_data,y_data,test_size=0.33)
     return tr_pairs,te_pairs,tr_y,te_y
+
+
+def speech_pairs_data(audio_group):
+    audio_samples = pd.read_csv('./outputs/'+audio_group+'.csv',names=['word','voice','rate','variant','file'])
+    audio_samples.loc[:,'file'] = audio_samples.loc[:,'file'].apply(lambda x:'outputs/'+audio_group+'/'+x).apply(generate_aiff_spectrogram)
+    y_data = audio_samples['variant'].apply(lambda x:x=='normal').values
+    max_samples = audio_samples['file'].apply(lambda x:x.shape[0]).max()
+    sample_size = audio_samples['file'][0].shape[1]
+    audio_samples_pos = audio_samples[audio_samples['variant'] == 'normal'].reset_index(drop=True)
+    audio_samples_neg = audio_samples[audio_samples['variant'] == 'phoneme'].reset_index(drop=True)
+    def append_zeros(spgr):
+        return np.lib.pad(spgr,[(0, max_samples-spgr.shape[0]), (0,0)],'median')
+    def create_data(sf):
+        sample_count = sf['file'].shape[0]
+        pad_sun = sf['file'].apply(append_zeros).values
+        x_data = np.vstack(pad_sun).reshape((sample_count,max_samples,sample_size))
+        return x_data
+    x_data_pos = create_data(audio_samples_pos)
+    x_data_neg = create_data(audio_samples_neg)
+    x_pos_train, x_pos_test, x_neg_train, x_neg_test =train_test_split(x_data_pos,x_data_neg,test_size=0.33)
+    tr_y = np.array(x_pos_train.shape[0]*[[1,0]])
+    te_y = np.array(x_pos_test.shape[0]*[[1,0]])
+    tr_pairs = np.array([x_pos_train,x_neg_train]).reshape(x_pos_train.shape[0],2,max_samples,sample_size)
+    te_pairs = np.array([x_pos_test,x_neg_test]).reshape(x_pos_test.shape[0],2,max_samples,sample_size)
+    return tr_pairs,te_pairs,tr_y,te_y
+
+if __name__ == '__main__':
+    print(speech_pairs_data())