unexpected (nan)reglin failure because of nan and Inf
Bug Description
nanreglin complains about nan in argument, strange
Steps to reproduce
with the x, y data in attachmentbgnanreglin.sod
--> load("TMP\bgnanreglin.sod")
ans =
T
--> [a,b]= nanreglin(x,y);
à la ligne 20 de la fonction reglin ( C:\Users\dc227480\AppData\Local\scilab-2024.0.0\modules\statistics\macros\reglin.sci ligne 32 )
à la ligne 37 de la fonction nanreglin ( C:\Users\dc227480\AppData\Local\scilab-2024.0.0\modules\statistics\macros\nanreglin.sci ligne 49 )
reglin : NaNs détectés, veuillez utiliser nanreglin() à la place.
What is the expected correct behavior?
remove the %nan points and perform the linear regression with reglin I suppose or explain better why reglin fails : actually because of %Inf in x vector which is not clearly catched and explained by reglin.
Edited by David Chèze