Checking Seasonality of a Time Series in Darts
Question:
I am currently working with my own dataset consisting of traffic volumes recorded at 5 minute intervals. I have data from 2018 to 2020.
I am working on some analysis of the data and I was looking through the Darts documentation here: https://unit8co.github.io/darts/generated_api/darts.utils.statistics.html
I am trying to check for seasonality utilizing the built in check_seasonality function they have however after reading through the documentation I am still somewhat confused on what my input for m needs to be? Would an integer input of m=1 mean it is checking for seasonality within 1 day? 1 month? 1 year?
Answers:
As mentioned in the documentation,
Checks whether the TimeSeries ts is seasonal with period m or not.
If m is None, we work under the assumption that there is a unique
seasonality period, which is inferred from the Auto-correlation
Function (ACF).
So, ‘m=1’ is the smallest unit time period according to your data set. It can be a day, month, or year.
I am currently working with my own dataset consisting of traffic volumes recorded at 5 minute intervals. I have data from 2018 to 2020.
I am working on some analysis of the data and I was looking through the Darts documentation here: https://unit8co.github.io/darts/generated_api/darts.utils.statistics.html
I am trying to check for seasonality utilizing the built in check_seasonality function they have however after reading through the documentation I am still somewhat confused on what my input for m needs to be? Would an integer input of m=1 mean it is checking for seasonality within 1 day? 1 month? 1 year?
As mentioned in the documentation,
Checks whether the TimeSeries ts is seasonal with period m or not.
If m is None, we work under the assumption that there is a unique
seasonality period, which is inferred from the Auto-correlation
Function (ACF).
So, ‘m=1’ is the smallest unit time period according to your data set. It can be a day, month, or year.