CODEWITHSUNDEEP
pythondataframetimetime-seriesdecomposition
Vipul Rustagi
Vipul Rustagi

Reputation: 85

None freq not allowing seasonal_decompose

Trying to apply seasonal_decompose on timeseries data whose freq is irregular. It looks something like this:

            modal_price
Period  
2014-11-01  1469
2015-01-01  1258
2015-03-01  1112
2015-04-01  1373
2015-06-01  1370
2015-07-01  1406
2015-08-01  1520
2015-09-01  1860
2015-10-01  1436
2015-11-01  1455

freq comes out to be None when I use df.index.freq

When I use seasonal_decompose function like this:

seasonal_decompose(x, model = 'additive')

it shows an error

ValueError: You must specify a freq or x must be a pandas object with a timeseries index with a freq not set to None.

Need help.

Upvotes: 3

Views: 4631

Answers (2)

Gonçalo Peres
Gonçalo Peres

Reputation: 13582

Statsmodel requires the frequency to decompose the series.

Usually, that frequency is in the metadata of the index (be it daily, weekly, mothly,...). However, yours doesn't have and that it why it gives you that error.

There at least two ways to solve it (let's consider the df named x):

• The one @A.Abs mentions, where you will pass freq in the seasonal_decompose, such as seasonal_decompose(x['Price'], freq=365).

• Setting the frequency to one's index column as x.index.adfreq(freq='d'), where 'd' corresponds to the day (daily frequency).

For more information on the frequency strings (or offset aliases) that can be passed into freq keyword arguments, check this answer or this Pandas Documentation page.

Upvotes: 0

A.Abs
A.Abs

Reputation: 470

I have faced the same problem and fixed it by specifying the frequency argument.

seasonal_decompose(Ts, model = 'additive', freq=1)

I hope this help. I found https://www.analyticsvidhya.com/blog/2016/02/time-series-forecasting-codes-python/ helpful.

Upvotes: 4

Related Questions