search

robjhyndman / forecast

Forecasting Functions for Time Series and Linear Models
http://pkg.robjhyndman.com/forecast
1.13k stars 342 forks source link

Larg and negative values in forecasting positive time series #141

Closed lassaw closed 9 years ago

lassaw commented 9 years ago

Thank you for your reply. here is an example of the bug i get :

" I used to forecast daily positive time series using tbats. But in some cases i get négatives or very long values like 1.25868e+14 where my real values are between 0 and 50000. I tried to include lamda=0 in my models but it doesn't work any more. I would like to know if there is a solution to avoid that problems ?"

Exemple of table datas

Code used r_code

Calling town_fc("lille",10) or town_fc("bayonne",10) These are 10 forecast values i get : forecast values

Thank you in advance.

Alassane

robjhyndman commented 9 years ago

Please provide a reproducible example of the problem. Also make sure you are using the latest version of the package as a similar problem to this was resolved in recent updates.

lassaw commented 9 years ago

Thank you sir, I don't get bug reports but just incorrect values in the output table. I'm updating the forecast package, now i have the 6.1 version, is it the last one ? Also, i notice that, the execution of tbats command takes lot of time, is there a way to reduce it. I have around 1500 observations for each country with many zeros ?

Thank you for your help

robjhyndman commented 9 years ago

This is not a help site. It is for bug reports only. Unless you provide a reproducible example of any problem, I will close this issue.

lassaw commented 9 years ago

Well received, i will manage to provide a reproducible example of any bug reports.