mlr-org mlr3 issues - Githubissues

mlr-org / mlr3

mlr3: Machine Learning in R - next generation

https://mlr3.mlr-org.com

GNU Lesser General Public License v3.0

947 stars 85 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

document Learner$hash, also do this for task and resampling

#1164 berndbischl closed 2 weeks ago
2
document Learner$hash, also do this for task and resampling

#1163 berndbischl closed 2 months ago
0
Converters for Resampling <-> data.table

#1162 mllg opened 2 months ago
5
Measure `"time_train"` does currently not take the time for the fallback learner into account

#1161 sebffischer closed 2 weeks ago
3
fix: empty predict set in score

#1160 be-marc closed 2 months ago
0
fix(prediction): predict_time for no predictions

#1159 sebffischer closed 2 months ago
0
Feature Request: Out-Of-Bag Error should not require storing the model.

#1158 sebffischer opened 2 months ago
0
fix: make rsq regression measure

#1157 be-marc closed 2 months ago
0
Error in private$.get_values() : when save learner in Linux, and load learner in Windows

#1156 lshddd closed 2 months ago
3
BREAKING CHANGE: encapsulation is not automatically activated when a fallback is set

#1155 be-marc closed 2 months ago
2
docs: update with new mlr3measure version

#1154 be-marc closed 2 months ago
0
feat: check column roles when using set_col_roles and col_roles

#1153 be-marc closed 2 months ago
0
encapsulate: doc what is in model slot, if encaps is on, but we fail due to an error

#1152 berndbischl closed 2 weeks ago
1
feat: default_fallback function

#1151 be-marc closed 2 months ago
0
have a nice function to create a default fallback -- and warn if it doesnt have enough properties

#1150 berndbischl closed 2 months ago
2
check whether we wanna allow assigment on task (measure, etc) properties

#1149 berndbischl opened 2 months ago
1
make sure that all "col roles" are properly documented

#1148 berndbischl closed 2 weeks ago
0
autotests: errorhandling seems weird, dont encapsulate

#1147 berndbischl closed 2 months ago
1
check that we have unit tests for wrong data entry when col roles are set

#1146 berndbischl closed 2 months ago
5
if we add: "measure_weights". more measures can be weighted

#1145 berndbischl opened 2 months ago
0
add an autotest, that tests the new predict_type = "quantiles"

#1144 berndbischl closed 2 months ago
1
document the autotest system

#1143 berndbischl closed 2 months ago
3
FeaturelessLearner: if "learner_weights" is added, the FFL should support weights too

#1142 berndbischl opened 2 months ago
1
document the "reflections" and their sytem in the developer wiki

#1141 berndbischl closed 4 weeks ago
1
fix NEWS typos

#1140 berndbischl closed 2 months ago
1
measure names (in mlr3measures) would be "cleaner" if we would use "regr_mse" not "mse" for function names

#1139 berndbischl closed 2 months ago
0
we really need to check code for metrics in terms of "sample_weights" and normalization.

#1138 berndbischl opened 2 months ago
2
docs in mlr3measures still say "weighted measures get normalized"

#1137 berndbischl opened 2 months ago
1
docs in mlr3measures still say "weighted measures get normalized"

#1136 berndbischl closed 2 months ago
1
review docs and code of new measure "mu_auc" : i also cannot find it...

#1135 berndbischl closed 2 months ago
2
check that a basic unit test for predict_type = "quantiles" is in the autotest for other learners

#1134 berndbischl closed 2 days ago
1
feat: set blas threads to 1

#1133 be-marc closed 2 months ago
0
feat: add measure properties to as.data.table.DictionaryMeasure

#1132 be-marc closed 2 months ago
0
fix: check type of column roles

#1131 be-marc closed 2 months ago
0
tests: check if predictions retrieved with as.data.table and predictions method are equal

#1130 be-marc closed 2 months ago
0
fix: ease expect_resample_result hash check

#1129 be-marc closed 2 months ago
0
docs: autotest

#1128 be-marc closed 2 months ago
4
fix: warn instead of error when feature types or levels differ

#1127 be-marc closed 2 months ago
0
chore: use fallback id

#1126 be-marc closed 2 months ago
0
feat: regr.featureless quantile prediction

#1125 be-marc closed 2 months ago
1
feat: add learner, resampling and measure weights

#1124 berndbischl opened 2 months ago
1
feat: debug learners new functions "selected_features" and "importance"

#1123 berndbischl closed 2 months ago
0
add comments and a mini fix for fallback

#1122 berndbischl closed 2 months ago
0
fix: warn instead of error when feature types or levels differ

#1121 sebffischer closed 2 months ago
1
Class weights no longer work

#1120 larskotthoff closed 2 months ago
1
BREAKING CHANGE: loglik, AIC, BIC, feature-creep

#1119 berndbischl closed 1 week ago
7
feat: classif.debug, regr.debug: add methods "importance" and "selected_features"

#1118 berndbischl closed 3 months ago
0
we should maybe check in learner / autotest that weights_learner influences the fit / the predictions of the learner

#1117 berndbischl opened 3 months ago
3
task_check_col_roles was pretty much empty and too lenient

#1116 berndbischl closed 2 months ago
1
Error with graph learner with preprocessing

#1115 larskotthoff closed 2 months ago
8

Previous Next