Add a function randc() to randomly generate data frame with categorical data, which are alphabetic characters. Extra characters combinations would be generated when the 26 characters are used up. (If number is desired, just leave a comment, I can update it)
Update the Corruptor class to accept an extra attribute dtype with default value np.float, so the Corrupter class can generate dataset in other dtype, like np.string
Add test cases for randc() function. One for BadInputError test, second for testing if the number of categories in the dataset is desired, third for testing if the shape of the dataset is desired.
This pull request is for addressing #67
dtype
with default valuenp.float
, so the Corrupter class can generate dataset in other dtype, like np.string