mxErrorPool() is a function that allows the user to retrieve a model that has been executed and returned with an error. If
mxRun() throws an error, then no model is returned to the user. This has the advantage of strongly reminding the user that the result of the optimization is very likely incorrect. However, this behavior is discouraging when debugging a model. One alternative is to use
mxRun(unsafe=TRUE) that will convert any errors into warnings. Another alternative is to retrieve the erroneous model using the mxErrorPool() function. The argument to the function is a character vector with the name of the model. Recall that the name of the model is specified by the
name= argument to mxModel() and is not necessarily the name of the variable that stores the model. If no argument is given to
mxErrorPool() then all erroneous models are returned in a list.