CODEWITHSUNDEEP
pythonpandasnumpydataframeautomation
Harish reddy
Harish reddy

Reputation: 431

Identifying only numeric values from a column in a Data Frame- Python

I Want a Separate column which returns "Yes" if the column "ID" contains all numeric values and 'No' if it contains alphabets or alphanumeric values.

ID      Result
3965      Yes
wyq8      No
RO_123    No
CMD_      No
2976      Yes

Upvotes: 1

Views: 2515

Answers (3)

Carmoreno
Carmoreno

Reputation: 1319

You can use .isnumeric() method:

df3["Result"] = df3["ID"].str.isnumeric().apply(lambda x: "No" if x == False else "Yes")

[UPDATE]: This method works only with integers numbers, please view the @Ch3steR answer for other cases.

Upvotes: 1

Ch3steR
Ch3steR

Reputation: 20669

You can use pd.Series.str.isnumeric here.

df['Result'] = np.where(df['ID'].str.isnumeric(), 'YES', 'NO')

       ID Result
0    3965    YES
1    wyq8     NO
2  RO_123     NO
3    CMD_     NO
4    2976    YES

There's a caveat with using isnumeric it doesn't identify float numbers.

test = pd.Series(["9.0", "9"])
test.str.isnumeric()

0    False
1     True
dtype: bool

If you strictly mark YES for int then use isnumeric else you can use pd.Series.str.fullmatch(available from version 1.1.0) here.

df['Result'] = np.where(df['ID'].str.fullmatch(r"\d+|\d+\.\d+", 'YES', 'NO')

For version <1.1.0 you use re.fullmatch

numeric_pat = re.compile(r"\d+|\d+\.\d+")
def numeric(val):
    match = numeric_pat.fullmatch(val)
    if match: return 'YES'
    else: return 'NO'

df['Result'] = df['ID'].apply(numeric)

Or we can use pd.to_numeric with boolean masking using pd.Series.isna

m = pd.to_numeric(df['ID'], errors='coerce').isna()
df['Result'] = np.where(m, 'NO', 'YES')

With errors parameter set to 'coerce' values which cannot be turned into numeic value will set to Nan.

test = pd.Series(['3965', 'wyq8', 'RO_123', 'CMD_', '2976'])
pd.to_numeric(test)

0    3965.0
1       NaN
2       NaN
3       NaN
4    2976.0
Name: ID, dtype: float64

Or you can build a custom function

def numeric(val):
    try:
        float(val)     # Using just `float` would suffice as int can be 
        return 'YES'   # converted to `float` so both `int`
                       # and `float` wouldnot raise any error
    except ValueError:
        return 'NO'

df['Result'] = df['ID'].apply(numeric)

Note: float handles scientic notation too, float("1e6") -> 1000000.0.

test = pd.Series(['1e6', '1', 'a 10', '1E6'])
test.apply(numeric)

0    YES
1    YES
2     NO
3    YES
dtype: object

Upvotes: 3

wwnde
wwnde

Reputation: 26676

Check if ID contains non-digitsand reverse the Boolean selection using ~. Using np.where, allocate option

df['Result']=np.where(~df.ID.str.contains('(\D+)'),'Yes','N0')

     ID Result
0    3965    Yes
1    wyq8     N0
2  RO_123     N0
3    CMD_     N0
4    2976    Yes

As noted by @Cameron Riddell. You could also skip inverting the boolean and do the following;

df['Result']=np.where(df.ID.str.contains('(\D+)'),'No','Yes')

Upvotes: 3

Related Questions