FR: Compact "how to load dirty data" example

[salamanders]

arff format is great! But for those of us dealing with crappy inputs, it might be helpful to have an example:

• For a CSV file
• With a header row
• And some extra lines at the end
• And some blanks
• And some Strings
• And some of the Strings are "T/F" and "Y/N"
• And some strings need splitting

What is the best (shortest?) path to a working RandomForest?

[haifengl]

There are test cases using csv files. you may search csv in code base.