zygmuntz / phraug Goto Github PK

View Code? Open in Web Editor NEW

271.0 20.0 152.0 322 KB

A set of simple Python scripts for pre-processing large files

Home Page: http://fastml.com/introducing-phraug/

License: BSD 2-Clause "Simplified" License

Python 100.00%

phraug's Issues

Not working

I did something like this:

#convert to libsvm format https://github.com/zygmuntz/phraug/blob/master/csv2libsvm.py
#python csv2libsvm.py spambase_renamed_class.csv libsvm.data -1 False

data = open("spambase/libsvm.data", "r")
data2 = open("spambase/libsvm2.data", "a")

i = 0
for line in data:
    temp = line[-3:]
    white_space_index = line[2:].find(" ")
    
    if i == 0:
        white_space_index = 0
        i += 1
    
    if "-1" in temp:
        line = "-1 " + line[white_space_index + 2:-6]
    else:
        line = "1 " + line[white_space_index + 2:-6]

        #drop last index (class) and first because it is the index nr.
    data2.write(line + "\n")
data.close()
data2.close()

preprocessing using libsvm format data

hi, I am newbie to spark, and want to using libsvm format data to do some feautre statistics and feature selection, now I have loaded libsvm data using MLUtils.loadLibSVMFile , but I don't know how to do next, do u have some suggestions about how to using libsvm data do feature statistics and selection? thx :)

Recommend Projects