convert a column type 'object' to float

Question

I have a column of numbers (type 'object') in my df:

1

2,23

5,23

200

I tried to convert using astype, but an error tells me that they cannot convert with commas, so I tried using

df[column].str.replace(',','.', inplace=True)

but the results tranformed all values without comma into NaNs..

So I tried another solution:

mask=df[column].str.contains(',',na=false) 
df.loc[mask].str.replace(',','.',inplace=True)

But an error message tells me that a dataframe has no str function, what could I do?

Answer 1

I have tried and df[column] = df[column].str.replace(',','.').astype(float) works fine for me.

Answer 2

I've found a solution for share:

mask=df[column].str.contains(',',na=False)
df[column]=df[column].mask(mask,df[column].str.replace(',','.'))