CODEWITHSUNDEEP
pythonpandasindexingdigits
user9187374
user9187374

Reputation: 309

python pandas get first digits of column values

I need to get for each row of a daframe, the first two digits of a number stored in another index (or column after a reset_index()). How can I do it?

My dataframe:

        value
index1       
110202      1
223168      5
850484      2
298008      3
950000      6
113500      6
849464      2
849616     10

I would like to obtain i.e:

                  value
index1 new_value       
110202 11             1
223168 22             5
850484 85             2
298008 29             3
950000 95             6
113500 11             6
849464 84             2
849616 84            10

Upvotes: 7

Views: 27295

Answers (2)

Dani Mesejo
Dani Mesejo

Reputation: 61910

Assuming index1 is the index of df you could do:

df['new_value'] = df.index.astype(str).str[:2]
print(df)

Output

        value new_value
index1                 
110202      1        11
223168      5        22
850484      2        85
298008      3        29
950000      6        95
113500      6        11
849464      2        84
849616     10        84

Basically convert the column to a string column and then use the str accessor to grab the first two characters. For more on working with text data, see here.

As an alternative you could reset the index and access the index1 column, for example:

df = df.reset_index()
df['new_value'] = df['index1'].astype(str).str[:2]
print(df.set_index(['index1', 'new_value']))

Output

                  value
index1 new_value       
110202 11             1
223168 22             5
850484 85             2
298008 29             3
950000 95             6
113500 11             6
849464 84             2
849616 84            10

Notice that in this alternative solution I set the index as the columns new_value and index1.

Upvotes: 11

Chrisvdberge
Chrisvdberge

Reputation: 1956

make a list from the index with df.index.values then iterate over the values in this array and grab the first 2 characters

Upvotes: 1

Related Questions