Implement converters parameter in read_csv()

[akharche]

Implement lambda converters in read_csv()

def category_converter(x):
      if x is '':
          return np.int32(0)
      else:
          return np.int32(int(x, 16))

names = ["{0}".format(i) for i in range(40)]
converter = {name: category_converter for name in names}

df = pd.read_csv(dataset_path, delimiter='\t', names=names, converters=converter)