>>> df.astype({'col1': 'int32'}).dtypes col1 int32 col2 int64 dtype: object
df = pd.read_csv("weather.tsv", sep="\t", dtype={'Day': str,'Wind':int64}) df.dtypes
