bug in plotPartialDependence for categorical predictors ?
2 次查看(过去 30 天)
显示 更早的评论
Hello
I am using regression tree with two predictors : one real and one categorical (10 categories from 0 to 9). When I plot the partial dependence for the categorical predictors, the estimates are given for 20 categories instead of 10 (at 0, 0.5, 1 etc. till 9 then NaN). I know that we can not use "query point" option to change these values but I do not understand how a partial dependence can be computed for a category that does not exist (as 0.5 or 1.5). Is it a bug ? My matlab version is 2018a and my script is
T=fitrtree(MODEL(train,2:npred),MODEL(train,1),'CategoricalPredictors',2);
plotPartialDependence(T,2);
[X,Y]=extract_fig2D(gcf); % this code simply extracts graphical values from a 2D plot
>> X
X =
Columns 1 through 14
0 0.5000 1.0000 1.5000 2.0000 2.5000 3.0000 3.5000 4.0000 4.5000 5.0000 5.5000 6.0000 6.5000
Columns 15 through 20
7.0000 7.5000 8.0000 8.5000 9.0000 NaN
Thank you in advance
Vincent M.
0 个评论
回答(0 个)
另请参阅
类别
在 Help Center 和 File Exchange 中查找有关 Linear Model Identification 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!