How to convert strings of different units to floats of the same?

July 6, 2024

I have a panda’s dataframe with strings, but I want floats. The units on the strings are different though. One column might be in mm with a few in cm. The next column might be g with a few in kg, and so on.

How can I convert each to floats in the same units? (i.e. if I’m changing everything to floats in mm, then "4.56 cm" should convert to 45.6, not 4.56).

>Solution :

Example

we need minimal and reproducible example to answer.

import pandas as pd
import numpy as np

data = {'length': ['4.56 cm', '100 mm', '3.5 cm', '10 mm'], 
        'weight': ['200 g', '0.5 kg', '300 g', '2 kg']}
df = pd.DataFrame(data)

df:

    length  weight
0  4.56 cm   200 g
1   100 mm  0.5 kg
2   3.5 cm   300 g
3    10 mm    2 kg

Code

I have prepared an answer with a example, but if your situation is different, please create and provide your own clear example.

s1 = df['length'].str.replace('[A-Za-z]', '', regex=True).astype('float')
df['length(mm)'] = np.where(df['length'].str.contains('cm'), s1 * 10, s1)

s2 = df['weight'].str.replace('[A-Za-z]', '', regex=True).astype('float')
df['weight(g)'] = np.where(df['weight'].str.contains('kg'), s2 * 1000, s2)

df:

    length  weight  length(mm)  weight(g)
0  4.56 cm   200 g        45.6      200.0
1   100 mm  0.5 kg       100.0      500.0
2   3.5 cm   300 g        35.0      300.0
3    10 mm    2 kg        10.0     2000.0