Skip to content

Python Cookbook

Handling Unnamed: n column in pandas

  • Set index to False when writing to csv file.
  • Set col index 0 when reading csv file.
pd.to_csv('flat_file.csv', index=False)   # set index to false when writing
pd.read_csv('flat_file.csv', index_col=0) # set index_col to 0 when reading

Change dtype in pandas

  • Types can be given at reading
  • After reading column type can be changed with to_numeric() astype() to_datetime()...

to_numeric() will change type to integer or float as appropriate. astype() let you specify the type you want. It accept a single type or dict of column names paired with types.

Reference

# Give column type when reading, setting dtypes of columns on dtype
df = pd.read_csv('flat_file.csv', dtype={'col_name': str}) 

# Changing types of dataframe column

# change col_name type to numeric. 
df.col_name = pd.to_numeric(df.col_name) 
# change col types with astype
df = df.astype({'col_name': str, 'col_name_2': float}) 
# changing single col
df.col_name = df.col_name.astype(float)