Error in data_ens[[2]] : subscript out of bounds about flexsdm

[KonstiDE]

How did you solve it? I have the same error, #338

[karolceron]

Did you solve it? Same error here, only for esm_max

[sjevelazco]

Hi all
Sorry, I forgot to work on this bug. I will work on this and let you know when it is fixed. Thanks

[sjevelazco]

Please,
Can one of you send a dataset to help me reproduce this error? A txt with presences (or presences/absences), .partition, and environmental data, and another .txt with background points with same columns. I can´t reproduce the error with the dataset provided by @dannyvelezv . Thanks

[KonstiDE]

Solved in #338, did this not work here? It was just an indexing problem that you could solve through just using more predictors

[sjevelazco]

It is not a good idea to double predictor to fix this problem. I will inspect those functions deeply. However, it will be more effective if one of you provides a dataset to reproduce this error. El sáb, 3 feb 2024 a las 5:56, Konstantin Müller ***@***.***>) escribió:

…

Solved in #338 <#338>, did this not work here? — Reply to this email directly, view it on GitHub <#318 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGW3YWAWPAIEFXKRURFYC4LYRX3VTAVCNFSM6AAAAAAQZFHWKSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRVGIZDQMJYGY> . You are receiving this because you modified the open/close state.Message ID: ***@***.***>

[sjevelazco]

The problem is you are using the esm_ family function with two predictors. It makes no sense because it is the same as constructing a single model with fit_ or tune_ family function. I added an error message to all esm_ functions. So, use tune or fit functions. Thank yo so much for reporting this error.
Best

[KonstiDE]

@sjevelazco, just have a look into this small project: https://github.com/KonstiDE/FlexSDM
Whilst running, just remove two of the predictor variables you dont like in the esm function family in flexsdm.R, Line 172 and following

[sjevelazco]

The code is ok. You can use esm_ with more than two. However, if you want to use esm_ with two variables, it will return an error because there is no ensemble of the bivariate models. In case you use two variables, use the fit_ family function. Best El lun, 5 feb 2024 a las 23:58, Konstantin Müller ***@***.***>) escribió:

…

@sjevelazco <https://github.com/sjevelazco>, just have a look into this small project: https://github.com/KonstiDE/FlexSDM Whilst running, just remove two of the predictor variables you dont like in the esm function family in fishi.R — Reply to this email directly, view it on GitHub <#318 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGW3YWDUVFWH7ZIJTY6KXJ3YSGL4DAVCNFSM6AAAAAAQZFHWKSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRYGY4TGMRWGI> . You are receiving this because you were mentioned.Message ID: ***@***.***>