On 16/04/19 14:37, Jamille Haarloo wrote:
Hi Moritz,
Thank you! it worked.
What worked, exactly ? ;-)
I did not find the line nor similar lines of 'features <-
na.omit(features)' in the v.class.mlR script/ R_script4 file.
Sorry, I am working with a heavily modified version here on my computer
currently, and didn't realize that this was part of my local modifications.
I also see that in the manual I actually wrote "The module makes no
effort to check the input data for NA values or anything else that might
perturb the analyses. It is up to the user to proceed to relevant checks
before launching the module."
I could add an na.omit to the code. What is your opinion on that as a
user ? Isn't it too invasive to just force this on the user ? I do
acknowledge that in my local case it is convenient.
Moritz
Best,
Jamille
On Mon, Apr 15, 2019 at 11:09 AM Moritz Lennert
<[email protected] <mailto:[email protected]>> wrote:
Hi Jamille,
On 15/04/19 14:49, Jamille Haarloo wrote:> Dear Moritz and other Grass-
users and developers,
>
> I tried dealing with the error myself by changing predicted <-
> data.frame(predict(models.cv <http://models.cv>
<http://models.cv>, features)) into
> predicted <- data.frame(predict(models.cv <http://models.cv>
<http://models.cv>, features,
> na.action = na.exclude)), based on discussions online implying some
> predictions might be invalid NaN values. I checked the script
output to
> see if this change was implemented and it was, but I get the
same error.
> Any suggestions what to try next?>
> ------------------------------
> v.class.mlR -i --overwrite segments_map=nvSegW24IDM4DV4@LUP1
> training_map=TrainingApril2019@LUP1 train_class_column=class_code
> output_class_column=output_class output_prob_column=probability
> classifiers=svmLinear,rf,xgbTree folds=5 partitions=10 tunelength=10
> weighting_modes=bwwv,qbwwv weighting_metric=accuracy
>
classification_results=C:\Users\haarlooj\Documents\CELOS\v.class.mlr_outputapril2019\results_all_classifiers
>
accuracy_file=C:\Users\haarlooj\Documents\CELOS\v.class.mlr_outputapril2019\accuracy_classifiers
>
model_details=C:\Users\haarlooj\Documents\CELOS\v.class.mlr_outputapril2019\details_classifier_module_runs
>
bw_plot_file=C:\Users\haarlooj\Documents\CELOS\v.class.mlr_outputapril2019\box-whicker_classifier_performance
>
r_script_file=C:\Users\haarlooj\Documents\CELOS\v.class.mlr_outputapril2019\R_script4
> processes=3
Normally, there should be no NA in the features as there is a line:
features <- na.omit(features)
early in the R script. Can you see it in the R_script4 file ?
> Running R now. Following output is R output.
> During startup - Warning messages:
> 1: Setting LC_CTYPE=en_US.cp1252 failed
> 2: Setting LC_COLLATE=en_US.cp1252 failed
> 3: Setting LC_TIME=en_US.cp1252 failed
> 4: Setting LC_MONETARY=en_US.cp1252 failed
> Loading required package: caret
> Loading required package: lattice
> Loading required package: ggplot2
> Warning messages:
> 1: package 'caret' was built under R version 3.5.3
> 2: package 'ggplot2' was built under R version 3.5.3
> Loading required package: foreach
> Loading required package: iterators
> Loading required package: parallel
> Warning messages:
> 1: package 'doParallel' was built under R version 3.5.3
> 2: package 'foreach' was built under R version 3.5.3
> 3: package 'iterators' was built under R version 3.5.3
> During startup - Warning messages:
> 1: Setting LC_CTYPE=en_US.cp1252 failed
> 2: Setting LC_COLLATE=en_US.cp1252 failed
> 3: Setting LC_TIME=en_US.cp1252 failed
> 4: Setting LC_MONETARY=en_US.cp1252 failed
> During startup - Warning messages:
> 1: Setting LC_CTYPE=en_US.cp1252 failed
> 2: Setting LC_COLLATE=en_US.cp1252 failed
> 3: Setting LC_TIME=en_US.cp1252 failed
> 4: Setting LC_MONETARY=en_US.cp1252 failed
> During startup - Warning messages:
> 1: Setting LC_CTYPE=en_US.cp1252 failed
> 2: Setting LC_COLLATE=en_US.cp1252 failed
> 3: Setting LC_TIME=en_US.cp1252 failed
> 4: Setting LC_MONETARY=en_US.cp1252 failed
> Error in data.frame(id = rownames(features), predicted) :
> arguments imply differing number of rows: 17851, 17849
> Execution halted
IDs are taken from the features and for some reasons there are two
features which do not have a prediction. It might help if you could
find
out why.
I cannot test right now, but you might want to check if you can replace
ids <- rownames(features)
with something like
ids <- rownames(predicted)
?
Moritz
_______________________________________________
grass-user mailing list
[email protected]
https://lists.osgeo.org/mailman/listinfo/grass-user
