[vsni.co.uk] Contact us
Author Message
Post new topic

  ASReml  ~  Bug with asreml R 4 - constraints mislabeled

kwstat
Posted: Tue Apr 21, 2020 7:12 pm Reply with quote
Joined: 18 Aug 2008 Posts: 25 Location: U.S.
In asreml R 3 the constraint column for "fa1" and "fa2" components was labeled as "U" for unconstrained. I believe this is correct.

Code:

 genotype:fa(site, 2)!site.BLA3071.fa1  0.2055    0.01283     16         U
 genotype:fa(site, 2)!site.MTA3071.fa1  0.2377    0.02049     12         U
 genotype:fa(site, 2)!site.PNA3071.fa1  0.1013    0.006714    15         U
 genotype:fa(site, 2)!site.RSA3071.fa1  0.2403    0.026        9.2       U
 genotype:fa(site, 2)!site.WTA3071.fa1  0.2273    0.0123      18         U
 genotype:fa(site, 2)!site.BLA3071.fa2  0                NA      NA      F
 genotype:fa(site, 2)!site.MTA3071.fa2  0.05263   0.02275      2.3       U
 genotype:fa(site, 2)!site.PNA3071.fa2 -0.006081  0.007743    -0.79      U
 genotype:fa(site, 2)!site.RSA3071.fa2  0.3043    0.01994     15         U
 genotype:fa(site, 2)!site.WTA3071.fa2 -0.03982   0.01505     -2.6       U


In asreml R 4 the bound column is labeled as "P" for positive. Shouldnt this be "U" instead of P? You can even see that the row for the 3rd fa2 component has a value of -.003199, which is negative, NOT positive.

Code:

 genotype:fa(site, 2)!BLA3071!fa1  0.2091    0.01289     16        P   0
 genotype:fa(site, 2)!MTA3071!fa1  0.232     0.0232      10        P   0
 genotype:fa(site, 2)!PNA3071!fa1  0.1032    0.006689    15        P   0
 genotype:fa(site, 2)!RSA3071!fa1  0.2113    0.02349      9        P   0
 genotype:fa(site, 2)!WTA3071!fa1  0.2226    0.01226     18        P   0
 genotype:fa(site, 2)!BLA3071!fa2  0                NA      NA     F  NA
 genotype:fa(site, 2)!MTA3071!fa2  0.2201    0.02428      9.1      P   0
 genotype:fa(site, 2)!PNA3071!fa2 -0.003199  0.009583    -0.33     P   0
 genotype:fa(site, 2)!RSA3071!fa2  0.11      0.03205      3.4      P   0
 genotype:fa(site, 2)!WTA3071!fa2  0.002106  0.01852      0.11     P   0


Possible fix: Change the bound labels for factor analytic terms from "P" to "U" (except for boundary/fixed labels).

_________________
Kevin Wright
View user's profile Send private message

Display posts from previous:  

All times are GMT
Page 1 of 1
Post new topic

Jump to:  

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You can attach files in this forum
You can download files in this forum