Looking for how to convert dataframe of numbers from objects to floats-CodePudding

I have the following dataframe: dataframe

I got the data by scrapping a basketball website but all the datatypes of the columns are objects instead of floats. I tried but splitting up the dataframe to only the numeric columns to do .astype(float) but I get an error. I just need help converting the proper columns into floats instead of objects.

Can't use .astype(float) cause the column has numbers that start with "." column example

How to remove the extra 0's screenshot

CodePudding user response：

This will allow you to change specific columns into a float

df[['column1', 'column2']] = df[['column1', 'column2']].astype(float)

CodePudding user response：

Use this pd.to_numeric with errors='ignore':

df.apply(pd.to_numeric, errors='ignore')

Here is a MCVE:

df = pd.DataFrame({'Letters':[*'ABCD'],
                   'Numbers':[1,2,3,4],
                   'Decimals':[.1,.2,.3,.4],
                   'String Numbers':['1', '.2', '1.5', '.04'],
                   'Words':'Dog Jane Tom Cat'.split(' ')})
    
df.apply(pd.to_numeric, errors='ignore')

Output:

  Letters  Numbers  Decimals  String Numbers Words
0       A        1       0.1            1.00   Dog
1       B        2       0.2            0.20  Jane
2       C        3       0.3            1.50   Tom
3       D        4       0.4            0.04   Cat

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 4 entries, 0 to 3
Data columns (total 5 columns):
 #   Column          Non-Null Count  Dtype  
---  ------          --------------  -----  
 0   Letters         4 non-null      object 
 1   Numbers         4 non-null      int64  
 2   Decimals        4 non-null      float64
 3   String Numbers  4 non-null      float64
 4   Words           4 non-null      object 
dtypes: float64(2), int64(1), object(2)
memory usage: 288.0  bytes