ENH: add an option to read_csv to select a specific rows / sample of the csv file

[jreback]

propose an option, sample or keeprows (#14285)

maybe taking a callable, for really just something like sample=10

to return you every 10th row

if it takes a callbale couuld use something like: lambda x: x % 10

easy way to get a sample of the csv file

useful to avoid specifying skiprows with a big list

and to solve an issue like this:

http://stackoverflow.com/questions/15555005/get-inferred-dataframe-types-iteratively-using-chunksize

[seven7e]

Oh, this issue has been opened for more than 3 years, and still opening? Why

[jorisvandenbossche]

Oh, this issue has been opened for more than 3 years, and still opening? Why

Because nobody has implemented it. Pull requests with code are certainly welcome.

[seven7e]

I would like to have a try, though it's "Difficulty Intermediate" and I am a newbie for the development of Pandas.

[linebp]

This seems to be already solved by this #15059. skiprows takes a callable with the linenumber as argument and returns true or false on whether or not to keep the row.