search

retowuest / autoMrP

autoMrP
25 stars 3 forks source link

EBMAforecast warning #5

Open philippbroniecki opened 4 years ago

philippbroniecki commented 4 years ago

In ebma.R, the following code leads to a warning message from the EBMAforecast package.

      EBMAforecast::makeForecastData(
        .predCalibration = data.frame(train_preds),
        .outcomeCalibration = as.numeric(unlist(train_y)),
        .predTest = data.frame(test_preds),
        .outcomeTest = as.numeric(unlist(test_y)))

Warning messages: 1: In if (class(.predCalibration) == "data.frame") { : the condition has length > 1 and only the first element will be used 2: In if (class(.predTest) == "data.frame") { : the condition has length > 1 and only the first element will be used 3: In if (class(.predCalibration) == "matrix") { : the condition has length > 1 and only the first element will be used 4: In if (class(.predTest) == "matrix") { : the condition has length > 1 and only the first element will be used

levi-polsci commented 2 years ago

I get a similar warning message in the serialize() function that is called within the ebma_mc_draws() function. The warning message I receive is: 

Warning in if (class(InvFisher) == "try-error") { :
  the condition has length > 1 and only the first element will be used
Warning in if (class(InvFisher) == "try-error") InvFisher <- try(solve(F_gross),  :
  the condition has length > 1 and only the first element will be used
Warning in if (class(InvFisher) == "try-error") { :
  the condition has length > 1 and only the first element will be used
Warning in if (s > 1) warning("Random slopes are not standardized back!") :
  the condition has length > 1 and only the first element will be used

I don't know if these warning are related but maybe they are...