search

JanDitzen / xtbreak

Testing and Estimation of structural breaks in Stata
https://janditzen.github.io/xtbreak/
18 stars 10 forks source link

xtbreak in stata17 takes too long to process #10

Open dayteri opened 2 years ago

dayteri commented 2 years ago

I am running xtbreak in stata17 on Mac but the xtbreak commands are taking a very long time to process (>3h) and hence I have to break before they complete. Is this a known problem? The data set I am using is very simple with one time variable as independent variable and a dependent variable, approx. 4000 observations.

lilijiang123 commented 4 months ago

I met the same problem. When I run xtbreak in stata16 for my daily data, approx. 4800 observations. xtbreak myvar, breakconstant

JanDitzen commented 4 months ago

Hi, can you please provide some more information? Is this a panel or a time series model, how large are the number of observations over time and space.

Especially for large time series, xtbreak calculates a large number of different possible solutions and therefore speed can be slow.

lilijiang123 commented 4 months ago

Hi, can you please provide some more information? Is this a panel or a time series model, how large are the number of observations over time and space.

Especially for large time series, xtbreak calculates a large number of different possible solutions and therefore speed can be slow.

Thank you very much for your response. This is a time series model, for daily volatility time series. The sample includes 4860 observations. When I used the "estat sbsingle" command, it didn't take much time. However, to my knowledge, "estat sbsingle" does not provide multiple breaks.

When I use monthly time series data, "xtbreak" runs smoothly.