Specifying number of decimal places for individual fields in pandas to csv

Question

I am using pandas to_csv function, and want to specify the number of decimal places for float numbers. However, I want this to change based on the field.

For example:

my data set looks like this:

sampleID     Call     X      Y 
1234         0.1234   0.123  0.123

I want the Call column to always have 4 decimal places and X and Y to always have 3.

Originally, I would use the float_format argument in to_csv, but that only appears applicable if all floats are treated the same way. How would I go about specifying the number of digits for individual columns?

Answer 1

You can round the columns before saving as a CSV. If you need the precision you can copy the DataFrame to retain precision in the original DataFrame.

df['X'] = df['X'].round(3)
df['Y'] = df['Y'].round(3)
df.to_csv('file.csv')