Convert dataframe with 2 columns into series-CodePudding

I have a dataframe named df with index being the date and im trying to convert it into a Series.

date                price      sales
2012-01-22          4200         0

I tried

pd.Series(df['price', 'sales'], index=df.index)

but it returns an error. May I know how can I overcome it?

CodePudding user response：

try df[['price', 'sales']].iloc[0].

code:

import pandas as pd
df = pd.DataFrame([{'date': pd.Timestamp('2012-01-22'), 'price':4200, 'sales': 0}])
df.set_index('date', inplace=True)
print(df)

series = df[['price', 'sales']].iloc[0]
print(series)

output:

            price  sales
date
2012-01-22   4200      0
price    4200
sales       0
Name: 2012-01-22 00:00:00, dtype: int64

CodePudding user response：

If you have a single row and want to convert to series you can use squeeze:

s = df.squeeze() # or df.squeeze(0)

Output:

price    4200
sales       0
Name: 2012-01-22 00:00:00, dtype: int64

Else, many options:

df.iloc[0]

df.loc['2012-01-22']

df.loc['2012-01-22', [['price', 'sales']]]

df.loc[df.index[0], [['price', 'sales']]]