

In [47]:

    
import pandas as pd

from os import listdir



In [2]:

    
!ls ../raw_data/









    



driver_imgs_list.csv  imgs                  sample_submission.csv

Data Summary

1) Sample Submission

Sum(c0:c9) == 1
Represent the predicted likelihood of each class
Expected Samples: 79,726



In [40]:

    
# Sample Data Raw
sample_df = pd.read_csv('../raw_data/sample_submission.csv')
print len(sample_df)
sample_df.head(1)



In [38]:

    
col_map = {
    'c0' : 'safe driving',
    'c1' : 'texting - right',
    'c2' : 'talking on the phone - right',
    'c3': 'texting - left',
    'c4': 'talking on the phone - left',
    'c5': 'operating the radio',
    'c6': 'drinking',
    'c7': 'reaching behind',
    'c8': 'hair and makeup',
    'c9': 'talking to passenger'
    }
sample_df.rename(columns=col_map).head(1)









    Out[38]:






  
    
      
      img
      safe driving
      texting - right
      talking on the phone - right
      texting - left
      talking on the phone - left
      operating the radio
      drinking
      reaching behind
      hair and makeup
      talking to passenger
    
  
  
    
      0
      img_1.jpg
      0.1
      0.1
      0.1
      0.1
      0.1
      0.1
      0.1
      0.1
      0.1
      0.1

2) Driver Images List

Training Dataset with proper classifications of some images
Train Size: 22,424



In [45]:

    
img_list_df = pd.read_csv('../raw_data/driver_imgs_list.csv')
print len(img_list_df)
img_list_df.head(1)









    



22424






    Out[45]:






  
    
      
      subject
      classname
      img
    
  
  
    
      0
      p002
      c0
      img_44733.jpg



In [ ]:

3) Image Directory

Contains a train/ and test/ directory



In [56]:

    
!ls ../raw_data/imgs/









    



test  train

Train Directory:

Contains a Direcotry for c0 through c9 classes (10 directories)

Class Directory Summary

c0 :: safe driving :: Length: 2490
c1 :: texting - right :: Length: 2267
c2 :: talking on the phone - right :: Length: 2317
c3 :: texting - left :: Length: 2346
c4 :: talking on the phone - left :: Length: 2326
c5 :: operating the radio :: Length: 2312
c6 :: drinking :: Length: 2325
c7 :: reaching behind :: Length: 2002
c8 :: hair and makeup :: Length: 1911
c9 :: talking to passenger :: Length: 2129



In [62]:

    
train_path = '../raw_data/imgs/train/'
train_files = listdir(train_path)
print len(train_files)
train_files









    



10






    Out[62]:





['c0', 'c1', 'c2', 'c3', 'c4', 'c5', 'c6', 'c7', 'c8', 'c9']



In [105]:

    
train_file_df = pd.DataFrame(columns=['file_path', 'class'])

for clas in train_files:
    class_files = listdir(train_path+clas+'/')
    if '.DS_Store' in class_files:
        class_files.remove('.DS_Store')
    # Create Dataframe with all files needed
    train_file_df = train_file_df.append(
        pd.DataFrame(zip([train_path+clas+'/'+f for f in class_files], [clas for _ in xrange(len(class_files))])
                    , columns=['file_path', 'class'])
                         )



In [108]:

    
train_file_df.head(2)









    Out[108]:






  
    
      
      file_path
      class
    
  
  
    
      0
      ../raw_data/imgs/train/c0/img_100026.jpg
      c0
    
    
      1
      ../raw_data/imgs/train/c0/img_10003.jpg
      c0



In [ ]:



In [ ]:

Test Directory

Contains a large number of images



In [ ]:

	file_path	class
0	../raw_data/imgs/train/c0/img_100026.jpg	c0
1	../raw_data/imgs/train/c0/img_10003.jpg	c0