CODEWITHSUNDEEP
pythonpandasdatetime
user3139545
user3139545

Reputation: 7394

Setting freq of pandas DatetimeIndex after DataFrame creation

Im using pandas datareader to get stock data. 

import pandas as pd
import pandas_datareader.data as web
ABB = web.DataReader(name='ABB.ST', 
                     data_source='yahoo',
                     start='2000-1-1')

However by default freq is not set on the resulting dataframe. I need freq to be able to navigate using the index like this:

for index, row in ABB.iterrows():
    ABB.loc[[index + 1]]

If freq is not set on DatetimeIndex im not able to use +1 etc to navigate.

What I have found are two functions astype and resample. Since I already know to freq resample looks like overkill, I just want to set freq to daily.

Now my question is how can i use astype on ABB to set freq to daily?

Upvotes: 36

Views: 68364

Answers (4)

Literalomicus
Literalomicus

Reputation: 33

Use .asfreq on the whole dataframe but it doesn't have an inplace argument so you need to save it an object.

ABB = ABB.asfreq("D")

Upvotes: 0

Rohit Nandi
Rohit Nandi

Reputation: 878

ABB is pandas DataFrame, whose index type is DatetimeIndex.

DatetimeIndex has freq attribute which can be set as below

ABB.index.freq = 'd'

Check out the change

ABB.index

Upvotes: 18

Abdou
Abdou

Reputation: 13284

Try:

ABB = ABB.asfreq('d')

This should change the frequency to daily with NaN for days without data.

Also, you should rewrite your for-loop as follows:

for index, row in ABB.iterrows():
    print(ABB.loc[[index + pd.Timedelta(days = 1)]])

Thanks!

Upvotes: 44

jezrael
jezrael

Reputation: 863501

If need change frequency of index resample is for you, but then need aggregate columns by some functions like mean or sum:

print (ABB.resample('d').mean())
print (ABB.resample('d').sum())

If need select another row use iloc with get_loc for find position of value in DatetimeIndex:

print (ABB.iloc[ABB.index.get_loc('2001-05-09') + 1])
Open            188.00
High            192.00
Low             187.00
Close           191.00
Volume       764200.00
Adj Close       184.31
Name: 2001-05-10 00:00:00, dtype: float64

Upvotes: 4

Related Questions