Closed Tato14 closed 2 years ago
Hi, trying to run your example I get:
R version 4.1.1 (2021-08-10)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19044)
Matrix products: default
locale:
[1] LC_COLLATE=English_United Kingdom.1252 LC_CTYPE=English_United Kingdom.1252
[3] LC_MONETARY=English_United Kingdom.1252 LC_NUMERIC=C
[5] LC_TIME=English_United Kingdom.1252
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] forcats_0.5.1 stringr_1.4.0 dplyr_1.0.7
[4] purrr_0.3.4 readr_2.0.2 tidyr_1.1.4
[7] tibble_3.1.6 ggplot2_3.3.5 tidyverse_1.3.1
[10] DALEXtra_2.1.1 DALEX_2.3.0 mlr3extralearners_0.5.18
[13] mlr3verse_0.2.2 mlr3pipelines_0.4.0 mlr3tuning_0.9.0
[16] paradox_0.7.1 mlr3learners_0.5.1 mlr3_0.13.0
@hbaniecki I am unsure why you cannot reproduce the code. In any case, it seems that the error is that I missed to encapsulate the graph as a learner with GraphLearner$new(graph)
command. This seems to solve the issue.
Transfered the issue, maybe we can improve the error messages here or check if we can do an auto-conversion.
@hbaniecki the problem here is probably that the value of myvar
depends on the result of list_mlr3learners()
, which in turn depends on your installed version of mlr3extralearners
(and to some degree on the versions of mlr3
and mlr3learners
that were installed on the machine where your mlr3extralearners
package was built).
You probably have a different myvar
than Tato14 has.
One problem is in predict.Graph
: we chose to emulate a lightweight Learner
here, but we always assume regression. Instead we should assume whatever the Graph
needs. However, ideally DALEXtra should auto-convert or at least check that its argument in explain_mlr3
is a Learner
(https://github.com/ModelOriented/DALEXtra/issues/71).
I created an ensemble model like:
This gives the following error:
I am unsure how to solve this, any idea? Thanks