Please refer the origin codes to here.
Extract the core component to convert .csv file into .ffm file, prepared for libffm
or xlearn
.
Please look up to run.py
and modify the source and destination path.
python run.py
./test.csv
contains 20 instances of 4 different categorical fields.
Tried to convert it into .ffm file in both parallel and serial way, and their time consumtion are as follow.