This shows you the differences between two versions of the page.
Next revision | Previous revision | ||
faq.parameter [2012/05/29 14:23] shogo created |
faq.parameter [2015/02/01 17:35] yutaka |
||
---|---|---|---|
Line 4: | Line 4: | ||
- Wrong positions or formats for observations and effects. Program may not stop and may get wrong results. Check outputs for the number of levels for each effect on the screen. | - Wrong positions or formats for observations and effects. Program may not stop and may get wrong results. Check outputs for the number of levels for each effect on the screen. | ||
- Missing or skipping one or more fixed lines in the parameter file. Program may stop. Check the missing line. | - Missing or skipping one or more fixed lines in the parameter file. Program may stop. Check the missing line. | ||
+ | - No extra empty line until OPTION or the end of the parameter file. | ||
- Misspelling. Program may stop. Correct the wrong spelling. | - Misspelling. Program may stop. Correct the wrong spelling. | ||
- | - Missing an empty last line. Program may not stop. Parameter, data, and pedigree files may need one more extra line at the end of the file. | + | - Missing an empty last (end) line. Programs may not stop. Parameter, data, and pedigree files may need one more extra line at the end of the file. |
- | - (Co)variance matrix is not symmetric, not positive definite, not right sized, ... Program may not stop. | + | - (Co)variance matrix is not symmetric, not positive definite, not right sized, ... Programs may not stop. |
- Scale down observations when using Gibbs sampling programs (e.g., 305-day milk yield 10000kg => 10000/1000 = 10). Program may not stop. | - Scale down observations when using Gibbs sampling programs (e.g., 305-day milk yield 10000kg => 10000/1000 = 10). Program may not stop. | ||
- A good result does not mean that your parameter file is correct. Always double-check! | - A good result does not mean that your parameter file is correct. Always double-check! | ||
+ | |||
+ | ====== covariance matrix for correlated effects in a multiple-trait model ====== | ||
+ | Example 1: two-trait model & direct/maternal effects (order of G is 4): | ||
+ | |||
+ | | ^Direct Trait1 ^Direct Trait2 ^Maternal Trait1 ^Maternal Trait 2 ^ | ||
+ | ^Direct Trait1 | | | | | | ||
+ | ^Direct Trait2 | | | | | | ||
+ | ^Maternal Trait1 | | | | | | ||
+ | ^Maternal Trait 2 | | | | | | ||
+ | |||
+ | |||
+ | Example 2: two-trait model & random regressions (RR) with 3 coefficients (order of G is 6): | ||
+ | |||
+ | | ^RR1 Trait1 ^RR1 Trait2 ^RR2 Trait1 ^RR2 Trait2 ^RR3 Trait1 ^RR3 Trait2 ^ | ||
+ | ^RR1 Trait1 | | | | | | | | ||
+ | ^RR1 Trait2 | | | | | | | | ||
+ | ^RR2 Trait1 | | | | | | | | ||
+ | ^RR2 Trait2 | | | | | | | | ||
+ | ^RR3 Trait1 | | | | | | | | ||
+ | ^RR3 Trait2 | | | | | | | |