singular matrix problem
-
- PLS Junior User
- Posts: 1
- Joined: Thu Sep 04, 2014 8:57 am
- Real name and title: tengku fauzan tengku anuar
singular matrix problem
i wonder why when i try to calculate PLS Algorithm the system appear singular matrix problem. could u please help me..
- Hengkov
- PLS Super-Expert
- Posts: 1599
- Joined: Sun Apr 24, 2011 10:13 am
- Real name and title: Hengky Latan
- Location: AMQ, Indonesia
- Contact:
Re: singular matrix problem
Hi,
Singularity Matrix Problem indicate you have HIGH COLLINEARITY PROBLEM or you have same case observation value.
For solve this problem, I recommend you add sample size.
Regards,
Singularity Matrix Problem indicate you have HIGH COLLINEARITY PROBLEM or you have same case observation value.
For solve this problem, I recommend you add sample size.
Regards,
Re: singular matrix problem
A relevant link for this discussion: http://www.talkstats.com/showthread.php ... e-function
I just resolved this issue. I encountered the singular matrix problem when attempting to bootstrap with a data set that I'd split by gender to test moderation. In my male data set, the data on 2 indicators were perfectly correlated. This is possible issue #2 on the reasons for singular matrix problems provided within PLS. To identify this issue, I ran a correlation matrix for all indicators in the model. When I identified those that were perfectly correlated, I dropped the one that had a slightly lower factor loading. I was then able to test my model.
I just resolved this issue. I encountered the singular matrix problem when attempting to bootstrap with a data set that I'd split by gender to test moderation. In my male data set, the data on 2 indicators were perfectly correlated. This is possible issue #2 on the reasons for singular matrix problems provided within PLS. To identify this issue, I ran a correlation matrix for all indicators in the model. When I identified those that were perfectly correlated, I dropped the one that had a slightly lower factor loading. I was then able to test my model.