This is an old revision of the document!
Table of Contents
FAQ: Error messages
Standard errors
I am getting a value of -1.0 as standard error for some of my estimates of genetic parameters, together with a note that Negative values for s.e. indicate failed approximation!. What does is mean and what can I do about it?
?
As it says in plain English: the approximation of standard errors failed.
Standard errors in WOMBAT are derived
- assuming large samples, and
- a series of approximations - see the Technical Details section of the manual.
In some cases, this approximation simply fails. Reasons for this may be, for instance, that your sample is very small, or that you are dealing with a model which is overparameterised. Please consult the statistical literature on maximum likelihood estimation for background information.
The latter includes multivariate analyses where some covariance matrices have eigenvalues which are effectively zero. In that scenario, it may help to fit a reduced rank model. Otherwise, you should attempt to use a `better' data set (i..e. one which supports the question you are asking) - if that is not feasible you may simply have to accept that the approximation of standard errors does not always work.
My REML analysis has converged but the results file does not report standard errors for (co)variance components or genetic parameters. How do I make the show up?
WOMBAT provides several algorithms to locate the maximum of the likelihood function. Only the `average information' algorithm provides an estimate of sampling covariances from the inverse of the information matrix (and s.e. derived from them). If you have used one of the other algorithms no estimates are thus reported - without further comments.
This means that you could try a continuation run with the average information algorithm - a single iterate should suffice to give you estimates of standard errors. Make sure though to inspect the likelihood - it should not change compared to the previous run: if it increases, your analysis did not reach convergence previously and you need to keep going, if it decreases something has gone wrong and results are not valid!
I have run a standard variance component analysis and found the solutions for effects fitted in the model. However, corresponding standard errors are missing - how do I get them?
In a mixed model analysis, approximate lower bound standard errors are obtained from the diagonal elements of the inverse of the coefficient matrix. The default method for variance component estimation is the “average informatiom” algorithm. The implementation in WOMBAT does not involve inversion of the coefficient matrix - hence standard errors are not simply a by-product.
You can enforce calculation of standard errors by simply adding the option FORCE-SE in a SPECIAL block at the end of the parameter file.
SPECIAL FORCE-SE END
Crash "Segmentation fault"
- Q: WOMBAT crashes with a “Segmentation fault” with a heap of unintelligible numbers to follow – this seems to happen quite early on, i.e. during the set-up phase before any likelihoods are printed to the screen. What is wrong?
- A: Long answer on separate page – here
Message "Small/Invalid Pivot"
- Q: When I run WOMBAT I sometimes get a message to the screen, saying either small, negative or invalid pivot - What does it mean and what should I do about it?
- A: Long answer on separate page – here
Message "lnkloc" : dimension exceeded !!
- Q: WOMBAT stops with this error message – why does this occur and, more importantly, is there anything I can do to make this go away (other than reducing the size of my analysis)?
- A: Depending on what exactly you are trying to do, this message can appear in different parts of the program – it is basically a 'safety catch' making sure your analysis has enough room. Most arrays needed are allocated at exactly the size required for the analysis; however, in some spots WOMBAT guesses at the space needed and, occasionally, this guess is simply wrong. In addition, there are some hard-coded maximum values and if your model of analysis is very big it may exceed one of these values.
- The first place this message may occur is during calculation of the inverse of the numerator relationship matrix (I have never seen this happen though); if it does, you are likely to have a very big and very dense pedigree and would be best advised to reduce the size of your analysis.
- The second possibility is that you are trying to use the PX-EM algorithm (this is the default for the first few iterates when there are more than 18 parameters to be estimated, so you may be doing this without having explicitly chosen to do so), and WOMBAT stops before the first iterate. WOMBAT has a guess on how many non-zero off-diagonal elements there are in one triangle of the coefficient matrix of the mixed model equations and then attempts to set this up and to determine the actual number. If this is too small, it will stop. A particular scenario when this has been found to occur it the case of an analysis trying to estimate a covariance matrix of reduced rank (option PC), where the rank is substantially less than the dimension of the matrix.
To get around this, you could try to:- use the AI-REML algorithm from the start (see run time option
—-aireml), especially if you have good starting values - set the initial guess manually using the run time option
—-choozhz
- The third place this can happen is when using the iterative solutions module; again, setting the maximum number manually via the
—-choozhzoption may help.
