Python pandas alternative to 'map' for function of 2 variables

Question

I would like to put 2 tables into 1, namely corporate announcement date and it's price on that day (on another table)

I have 2 dataframes with the following columns

df1: date, announcement, ticker

date        ticker   announcement
25/4/2013   AAPL     Change in Boardroom
25/4/2013   GOOG     OTHERS
25/4/2013   AMZN     Change in Audit Committee

df2: date, ticker, price

date        ticker   announcement
22/3/2012   AAPL     100.00
23/3/2012   AAPL     102.30
24/3/2012   AAPL     105.40
...
...

def getPrice(dt,tk):
    try:
        return df2[(df2['date']>=dt)&(df2['ticker']==tk)].sort_values(by='date')['price'].values[0]
    except:
        return 0

prices_array = list(map(getPrice,df1['date'].values,df1['ticker'].values))

df1['price'] = prices_array

For the "map" function, the whole process takes very long time. I would love to use apply for df, but I only know how to use apply with lambda function which does not contain "if, then".

I want an extra column like this:

date        ticker   announcement               price
25/4/2013   AAPL     Change in Boardroom        124.10
25/4/2013   GOOG     OTHERS                     50.85
25/4/2013   AMZN     Change in Audit Committee  102.20

Any suggestion on quick ways to do this? Or can do this in a short amount of time?

Thank you and appreciate your help

Can you share your dummy dataframe and also your expected output? It's easier to understand that way. — Mohit Motwani
– Mohit Motwani, Commented Dec 4, 2018 at 5:59

Mark · Accepted Answer · 2019-05-14 09:10:58Z

1

I think you can use '.join'

df.set_index(['date', 'ticker']).join(df2.set_index(['date', 'ticker'])

answered May 14, 2019 at 9:10

Mark

9841 gold badge11 silver badges27 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

gilch · Accepted Answer · 2018-12-04 06:04:47Z

0

You can create your function for DataFrame.apply() with def instead of lambda. Define it in advance and then put the function name where the lambda would go, same as you did for map().

The applied function gets a Series argument, so you can get the whole row.

edited Dec 4, 2018 at 6:04

answered Dec 4, 2018 at 5:57

gilch

12k3 gold badges25 silver badges28 bronze badges

1 Comment

lrh09 Over a year ago

can you answer the question instead of a generic one? Thanks

Collectives™ on Stack Overflow

Python pandas alternative to 'map' for function of 2 variables

2 Answers 2

Comments

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related