error "using reshape"

announcements (new versions, changes, bugs, installation problems...) related to M-SURGE, E-SURGE and U-CARE

error "using reshape"

Postby AlexAv » Wed Sep 19, 2012 1:23 pm

Hi Rémi,

I am trying to fit multi-event models with E-surge for Windows 64. It is a huge model with 25 initial states, 4 transition steps (among them, two transition matrices have dimension 25*41 and 41*25) and 3 event steps with 5 possible events.
The model is huge because I have included trap-aware states to account for trap-dependence in the data.
Hence, the matrix dealing with trapawareness is the last transition matrix as R.Pradel and A Sanz-AGuilar did recently.
Gepat, Gemaco (albeit long) work well. I use initial values (using IVFV files) from previous betas of simpler models (so I used real axis). the IVFV step works too, but, as soon as I click on the run button, It get an error message in the black window:

here the beginning of the message:
"Error using reshape
To RESHAPE the number of elements must not change.
Error in modme/grad_F1_age_evt>gradPattern (line 723)
Error in modme/grad_F1_age_evt (line 683)
Error in modme /f_age (line 19)
[...]
"

what does it mean?
Is this due to a too large matrix dimension?

Could you help me please?

Best wishes

Alexis
AlexAv
 
Posts: 12
Joined: Tue Sep 18, 2012 8:19 am
Location: Sweden and France

Re: error "using reshape"

Postby CHOQUET » Thu Sep 20, 2012 11:42 am

Hello Alexis,
did you try the model straigthly, i.e. without using initial values
from a previous model ?.
did you use initial values from a model with the same structure ?
Rémi
CHOQUET
 
Posts: 211
Joined: Thu Nov 24, 2005 4:58 am
Location: CEFE, Montpellier, FRANCE.

Re: error "using reshape"

Postby AlexAv » Thu Sep 20, 2012 1:20 pm

Hi Rémi,

yes, I tried the model without initial values and the error message remained. The IV were not from the same model structure (the IV came from the model without trap aware states), but I checked the correspondence between the betas/parameters.

Alexis
AlexAv
 
Posts: 12
Joined: Tue Sep 18, 2012 8:19 am
Location: Sweden and France

Re: error "using reshape"

Postby CHOQUET » Fri Sep 21, 2012 4:11 am

Can you check your pattern matrices ?
Only one '*' is needed and allowed by row. Otherwise, each row is composed of '-' and 'letters'.
CHOQUET
 
Posts: 211
Joined: Thu Nov 24, 2005 4:58 am
Location: CEFE, Montpellier, FRANCE.

Re: error "using reshape"

Postby AlexAv » Fri Sep 21, 2012 5:05 am

Hi Rémi,

thanks for your responses.
I have already checked the transition and the capture histories too. The matrices are row stochastic and I have one "*" per row. The model is "almost" simple as I have only one parameter per row. Furthermore, the total number of parameter (of the umbrella model) is weak :25.
Note that the number of capture occasions is more than 150.

Note also that I use E-Surge 1.8.5 with matlab libraries for windows 64 (my computer works on windows 64)

sincerly

Alexis
AlexAv
 
Posts: 12
Joined: Tue Sep 18, 2012 8:19 am
Location: Sweden and France

Re: error "using reshape"

Postby CHOQUET » Mon Sep 24, 2012 6:59 am

Hello Alexis,
I don't know what can happen.
Can you send me the data and the pattern files so that I can reproduce the error ?
Sincerely,
Rémi
CHOQUET
 
Posts: 211
Joined: Thu Nov 24, 2005 4:58 am
Location: CEFE, Montpellier, FRANCE.

Re: error "using reshape"

Postby AlexAv » Mon Oct 01, 2012 9:33 am

Hi Rémi,

I sent the data

thanks

Alexis
AlexAv
 
Posts: 12
Joined: Tue Sep 18, 2012 8:19 am
Location: Sweden and France


Return to software problems/news

Who is online

Users browsing this forum: No registered users and 1 guest