Convert data frame to numeric and dropna

Question

I been using:

df = df.convert_objects(convert_numeric=True).dropna()

but since it's being deprecated I need to replace it with to_numeric

I've tried

df = pd.to_numeric
df = df.dropna()

Sample data:

Name  Race Fav Age  Weight Height Style Cut
John   D    K  23    120   23.5    DD   RET
Rose   Z    U  33    110   47.9    KZ   DEZ
James  Z    U  FF    UK    NOT     Z    R

Would like to convert to drop the rows that are non numeric

Output:

Name  Race Fav Age  Weight Height Style Cut
John   D    K  23    120   23.5    DD   RET
Rose   Z    U  33    110   47.9    KZ   DEZ

it depends on the data also did you try https://stackoverflow.com/questions/15891038/change-data-type-of-columns-in-pandas? — EdChum, Nov 08 '17 at 16:58

score 1 · Accepted Answer · answered Nov 08 '17 at 17:27

1

I'd do it this way:

In [399]: num_cols = df.columns[df.apply(pd.to_numeric, errors='coerce').any()]

In [400]: df[num_cols] = df[num_cols].apply(pd.to_numeric, errors='coerce')

In [401]: df = df[df.select_dtypes(['number']).notnull().all(1)]

In [402]: df
Out[402]:
   Name Race Fav   Age  Weight  Height Style  Cut
0  John    D   K  23.0   120.0    23.5    DD  RET
1  Rose    Z   U  33.0   110.0    47.9    KZ  DEZ

In [403]: df.dtypes
Out[403]:
Name       object
Race       object
Fav        object
Age       float64
Weight    float64
Height    float64
Style      object
Cut        object
dtype: object

answered Nov 08 '17 at 17:27

MaxU - stand with Ukraine

205,989
36
386
419

I get this error:TypeError: ("to_numeric() got an unexpected keyword argument 'error'", 'occurred at index Age') – magicsword Nov 08 '17 at 17:34
1

Perfect. That Worked! Thanks. – magicsword Nov 08 '17 at 17:37

score 0 · Answer 2 · answered Nov 08 '17 at 17:02

0

You aren't calling to_numeric properly. Since it only applies to columns, if you want to apply it to all columns you'll have to use pd.apply

df = df.apply(pd.to_numeric)
df = df.dropna()

answered Nov 08 '17 at 17:02

Louise Davies

14,781
6
38
41

score 0 · Answer 3 · answered Nov 08 '17 at 17:15

0

apply the to_numeric then dropna

df.apply(lambda x :pd.to_numeric(x, errors ='coerce'),axis=1).dropna()

answered Nov 08 '17 at 17:15

BENY

317,841
20
164
234

Convert data frame to numeric and dropna

3 Answers3