speed up pandas calculation

Vincent Davis vincent at vincentdavis.net
Thu Jul 31 01:04:04 CEST 2014


I know this is a general python list and I am asking about pandas but this
question is probably not great for asking on stackoverflow.
I have a list of files (~80 files, ~30,000 rows) I need to process with my
current code it is take minutes for each file. Any suggestions of a fast
way. I am try to stick with pandas for educational purposes. Any
suggestions would be great. If you are curious the can find the data file I
am using below here. http://www.nber.org/nhamcs/data/nhamcsopd2010.csv

drugs_current = {'CITALOPRAM': 4332,
         'ESCITALOPRAM': 4812,
         'FLUOXETINE': 236,
         'FLUVOXAMINE': 3804,
         'PAROXETINE': 3157,
         'SERTRALINE': 880,
         'METHYLPHENIDATE': 900,
         'DEXMETHYLPHENIDATE': 4777,
         'AMPHETAMINE-DEXTROAMPHETAMINE': 4035,
         'DEXTROAMPHETAMINE': 804,
         'LISDEXAMFETAMINE': 6663,
         'METHAMPHETAMINE': 805,
         'ATOMOXETINE': 4827,
         'CLONIDINE': 44,
         'GUANFACINE': 717}

drugs_98_05 = { 'SERTRALINE': 56635,
                'CITALOPRAM': 59829,
                'FLUOXETINE': 80006,
                'PAROXETINE_HCL': 57150,
                'FLUVOXAMINE': 57064,
                'ESCITALOPRAM': 70466,
                'DEXMETHYLPHENIDATE': 70427,
                'METHYLPHENIDATE': 70374,
                'METHAMPHETAMINE': 53485,
                'AMPHETAMINE1': 70257,
                'AMPHETAMINE2': 70258,
                'AMPHETAMINE3': 50265,
                'DEXTROAMPHETAMINE1': 70259,
                'DEXTROAMPHETAMINE2': 70260,
                'DEXTROAMPHETAMINE3': 51665,
                'COMBINATION_PRODUCT': 51380,
                'FIXED_COMBINATION': 51381,
                'ATOMOXETINE': 70687,
                'CLONIDINE1': 51275,
                'CLONIDINE2': 70357,
                'GUANFACINE': 52498
               }

df = pd.read_csv('nhamcsopd2010.csv' , index_col='PATCODE',
low_memory=False)
col_init = list(df.columns.values)
keep_col = ['PATCODE', 'PATWT', 'VDAY', 'VMONTH', 'VYEAR', 'MED1', 'MED2',
'MED3', 'MED4', 'MED5']
for col in col_init:
    if col not in keep_col:
        del df[col]
if f[-3:] == 'csv' and f[-6:-4] in ('93', '94', '95', '96', '97', '98',
'99', '00', '91', '02', '03', '04', '05'):
    drugs = drugs_98_05
elif f[-3:]  == 'csv' and f[-6:-4] in ('06', '08', '09', '10'):
    drugs = drugs_current
for n in drugs:
    df[n] = df[['MED1','MED2','MED3','MED4','MED5']].isin([drugs[n]]).any(1)


Vincent Davis
720-301-3003
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20140730/3acff346/attachment.html>


More information about the Python-list mailing list