CODEWITHSUNDEEP
pythonpandasdataframe
Chuck
Chuck

Reputation: 4892

How do I convert two DataFrame columns into summed Series?

I have a pandas DataFrame that looks like this:

              date  sku  qty
0       2015-10-30  ABC    1
1       2015-10-30  DEF    1
2       2015-10-30  ABC    2
3       2015-10-31  DEF    1
4       2015-10-31  ABC    1
...            ...  ...  ...

How can extract all of the data for a particular sku and sum up the qty by date. For example, the ABC SKU?

2015-10-30   3
2015-10-31   1
       ... ...

The closest I've gotten is a hierarchal grouping with sales.groupby(['date', 'sku']).sum().

Upvotes: 0

Views: 29

Answers (1)

Quang Hoang
Quang Hoang

Reputation: 150745

If you will work with all (or several) sku, then:

agg_df = df.groupby(['sku','date']).qty.sum()

# extract some sku data
agg_df.loc['ABC']

Output:

date
2015-10-30    3
2015-10-31    1
Name: qty, dtype: int64

If you only care for ABC particularly, then it's better to filter it first

df[df['sku'].eq('ABC')].groupby('date')['qty'].sum()

The output would be the same as above.

Upvotes: 2

Related Questions