How to replace nan values in read_csv?

Question

I'm reading a flat file with read_csv and flat file as some NaN values because of which it is changing the datatype to float instead of integer which I dont want.

Is there any way to replace the NaN values with 0 in the read_csv method?

I know I can do df.fillna(value=0) but my question is how can I replace these nan values in the read_csv .

decop=pd.read_csv("C:\\Users\\mnk044\\Downloads\\amna_decp_part-m-00000", sep="\x01", names=clist)

Answer 1

Filling NaN values while loading data is currently not supported (see the read_csv docs for information on all the supported functionality).

Your only option is to first read, and then call fillna .

df = pd.read_csv(filepath)
df.fillna(0, inplace=True)

Next, if you want to convert your float columns to integers, then filter on df.dtypes .

c = df.columns[df.dtypes == float]
df[c] = df[c].astype(int)