help
86 Subscribers 
help > More ROIs, more problems
Showing 1-5 of 5 posts
Oct 22, 2014 08:10 PM | Simon Davis
More ROIs, more problems
Hiya,
I'm trying to adapt an NBS pipeline for a set of 471 ROIs; we ran the same analysis on a lower-dimension set (112 ROIs) and everything ran fine, but now i'm getting a "Stop: Design Matrix not found or inconsistent" error. I suspect an error with memory; if i reduce the number of subjects (from 47 to 27), NBS runs the model. Unfortunately i'm not sure how to debug to find out if this is truly a memory error or not...
Attached is one of the design matrices i've tried. (47 subjects, 2 conditions).
I'm trying to adapt an NBS pipeline for a set of 471 ROIs; we ran the same analysis on a lower-dimension set (112 ROIs) and everything ran fine, but now i'm getting a "Stop: Design Matrix not found or inconsistent" error. I suspect an error with memory; if i reduce the number of subjects (from 47 to 27), NBS runs the model. Unfortunately i'm not sure how to debug to find out if this is truly a memory error or not...
Attached is one of the design matrices i've tried. (47 subjects, 2 conditions).
Oct 22, 2014 08:10 PM | Simon Davis
RE: More ROIs, more problems
and a sample subject from this analysis
Oct 23, 2014 12:10 AM | Andrew Zalesky
RE: More ROIs, more problems
Hi Simon,
This could potentially be a memory problem, but I'd be surprised if that was the case.
The reason this error appears is typically due to the number of rows in the design matrix not exactly equaling the number of connectivity matrices.
Also, check that the contrast has as many values as there are columns in the design matrix.
It seems you have all this worked out though, given everything is working for the lower dimension data set.
If it is truly a memory issue, you might want to try on a system with more RAM.
If you send me your connectivity matrices and design matrix, I'm happy to run it here.
Andrew
Originally posted by Simon Davis:
This could potentially be a memory problem, but I'd be surprised if that was the case.
The reason this error appears is typically due to the number of rows in the design matrix not exactly equaling the number of connectivity matrices.
Also, check that the contrast has as many values as there are columns in the design matrix.
It seems you have all this worked out though, given everything is working for the lower dimension data set.
If it is truly a memory issue, you might want to try on a system with more RAM.
If you send me your connectivity matrices and design matrix, I'm happy to run it here.
Andrew
Originally posted by Simon Davis:
Hiya,
I'm trying to adapt an NBS pipeline for a set of 471 ROIs; we ran the same analysis on a lower-dimension set (112 ROIs) and everything ran fine, but now i'm getting a "Stop: Design Matrix not found or inconsistent" error. I suspect an error with memory; if i reduce the number of subjects (from 47 to 27), NBS runs the model. Unfortunately i'm not sure how to debug to find out if this is truly a memory error or not...
Attached is one of the design matrices i've tried. (47 subjects, 2 conditions).
I'm trying to adapt an NBS pipeline for a set of 471 ROIs; we ran the same analysis on a lower-dimension set (112 ROIs) and everything ran fine, but now i'm getting a "Stop: Design Matrix not found or inconsistent" error. I suspect an error with memory; if i reduce the number of subjects (from 47 to 27), NBS runs the model. Unfortunately i'm not sure how to debug to find out if this is truly a memory error or not...
Attached is one of the design matrices i've tried. (47 subjects, 2 conditions).
Oct 23, 2014 12:10 AM | Simon Davis
RE: More ROIs, more problems
Excellent, thanks! I've send a link via your gmail. Let me
know if you need and more info.
S
Originally posted by Andrew Zalesky:
S
Originally posted by Andrew Zalesky:
Hi
Simon,
This could potentially be a memory problem, but I'd be surprised if that was the case.
The reason this error appears is typically due to the number of rows in the design matrix not exactly equaling the number of connectivity matrices.
Also, check that the contrast has as many values as there are columns in the design matrix.
It seems you have all this worked out though, given everything is working for the lower dimension data set.
If it is truly a memory issue, you might want to try on a system with more RAM.
If you send me your connectivity matrices and design matrix, I'm happy to run it here.
Andrew
Originally posted by Simon Davis:
This could potentially be a memory problem, but I'd be surprised if that was the case.
The reason this error appears is typically due to the number of rows in the design matrix not exactly equaling the number of connectivity matrices.
Also, check that the contrast has as many values as there are columns in the design matrix.
It seems you have all this worked out though, given everything is working for the lower dimension data set.
If it is truly a memory issue, you might want to try on a system with more RAM.
If you send me your connectivity matrices and design matrix, I'm happy to run it here.
Andrew
Originally posted by Simon Davis:
Hiya,
I'm trying to adapt an NBS pipeline for a set of 471 ROIs; we ran the same analysis on a lower-dimension set (112 ROIs) and everything ran fine, but now i'm getting a "Stop: Design Matrix not found or inconsistent" error. I suspect an error with memory; if i reduce the number of subjects (from 47 to 27), NBS runs the model. Unfortunately i'm not sure how to debug to find out if this is truly a memory error or not...
Attached is one of the design matrices i've tried. (47 subjects, 2 conditions).
I'm trying to adapt an NBS pipeline for a set of 471 ROIs; we ran the same analysis on a lower-dimension set (112 ROIs) and everything ran fine, but now i'm getting a "Stop: Design Matrix not found or inconsistent" error. I suspect an error with memory; if i reduce the number of subjects (from 47 to 27), NBS runs the model. Unfortunately i'm not sure how to debug to find out if this is truly a memory error or not...
Attached is one of the design matrices i've tried. (47 subjects, 2 conditions).
Oct 24, 2014 01:10 AM | Andrew Zalesky
RE: More ROIs, more problems
Hi Simon, I've emailed you. Andrew
Originally posted by Simon Davis:
Originally posted by Simon Davis:
Excellent, thanks! I've send a link via
your gmail. Let me know if you need and more info.
S
Originally posted by Andrew Zalesky:
S
Originally posted by Andrew Zalesky:
Hi
Simon,
This could potentially be a memory problem, but I'd be surprised if that was the case.
The reason this error appears is typically due to the number of rows in the design matrix not exactly equaling the number of connectivity matrices.
Also, check that the contrast has as many values as there are columns in the design matrix.
It seems you have all this worked out though, given everything is working for the lower dimension data set.
If it is truly a memory issue, you might want to try on a system with more RAM.
If you send me your connectivity matrices and design matrix, I'm happy to run it here.
Andrew
Originally posted by Simon Davis:
This could potentially be a memory problem, but I'd be surprised if that was the case.
The reason this error appears is typically due to the number of rows in the design matrix not exactly equaling the number of connectivity matrices.
Also, check that the contrast has as many values as there are columns in the design matrix.
It seems you have all this worked out though, given everything is working for the lower dimension data set.
If it is truly a memory issue, you might want to try on a system with more RAM.
If you send me your connectivity matrices and design matrix, I'm happy to run it here.
Andrew
Originally posted by Simon Davis:
Hiya,
I'm trying to adapt an NBS pipeline for a set of 471 ROIs; we ran the same analysis on a lower-dimension set (112 ROIs) and everything ran fine, but now i'm getting a "Stop: Design Matrix not found or inconsistent" error. I suspect an error with memory; if i reduce the number of subjects (from 47 to 27), NBS runs the model. Unfortunately i'm not sure how to debug to find out if this is truly a memory error or not...
Attached is one of the design matrices i've tried. (47 subjects, 2 conditions).
I'm trying to adapt an NBS pipeline for a set of 471 ROIs; we ran the same analysis on a lower-dimension set (112 ROIs) and everything ran fine, but now i'm getting a "Stop: Design Matrix not found or inconsistent" error. I suspect an error with memory; if i reduce the number of subjects (from 47 to 27), NBS runs the model. Unfortunately i'm not sure how to debug to find out if this is truly a memory error or not...
Attached is one of the design matrices i've tried. (47 subjects, 2 conditions).