How to zip two columns into a key value pair dictionary in pandas

Question

I have a dataframe with two related columns that needs to be merged into a single dictionary column.

Sample Data:

    skuId   coreAttributes.price    coreAttributes.amount
0   100     price                   8.84
1   102     price                   12.99
2   103     price                   9.99

Expected output:

skuId    coreAttributes
100      {'price': 8.84}
102      {'price': 12.99}
103      {'price': 9.99}

What I’ve tried:

planProducts_T = planProducts.filter(regex = 'coreAttributes').T
planProducts_T.columns = planProducts_T.iloc[0]
planProducts_T.iloc[1:].to_dict(orient = 'records')

I get UserWarning: DataFrame columns are not unique, some columns will be omitted. and this output:

[{'price': 9.99}]

Could you someone please help me on this.

Asked By: Karthik S

||

Source

Answer 1

You can use a list comprehension with python’s zip:

df['coreAttributes'] = [{k: v} for k,v in
                        zip(df['coreAttributes.price'],
                            df['coreAttributes.amount'])]

Output:

   skuId coreAttributes.price  coreAttributes.amount    coreAttributes
0    100                price                   8.84   {'price': 8.84}
1    102                price                  12.99  {'price': 12.99}
2    103                price                   9.99   {'price': 9.99}

If you need to remove the initial columns, use pop.

df['coreAttributes'] = [{k: v} for k,v in
                        zip(df.pop('coreAttributes.price'),
                            df.pop('coreAttributes.amount'))]

Output:

   skuId    coreAttributes
0    100   {'price': 8.84}
1    102  {'price': 12.99}
2    103   {'price': 9.99}

Answered By: mozway

Answer 2

you can use apply and drop for an optimize computation

df["coreAttributes"] = df.apply(lambda row: {row["coreAttributes.price"]: row["coreAttributes.amount"]}, axis=1)
df.drop(["coreAttributes.price","coreAttributes.amount"], axis=1)

output

    skuId   coreAttributes
0   100     {'price': 8.84}
1   102     {'price': 12.99}
2   103     {'price': 9.99}

Answered By: Lucas M. Uriarte

Answer 3

df.set_index("skuId").apply(lambda ss:{ss[0]:ss[1]},axis=1).rename("coreAttributes").reset_index()

out：

 skuId    coreAttributes
0    100   {'price': 8.84}
1    102  {'price': 12.99}
2    103   {'price': 9.99}

Answered By: G.G

How to zip two columns into a key value pair dictionary in pandas

Question:

Answers: