Ggtcf

Need help identifying/translating a plaque in Tangier, Morocco

Can Medicine checks be used, with decent rolls, to completely mitigate the risk of death from ongoing damage?

Distance between two points on a map made for a game

What is the white spray-pattern residue inside these Falcon Heavy nozzles?

What is the meaning of "of trouble" in the following sentence?

"which" command doesn't work / path of Safari?

Why do we use polarized capacitors?

How can bays and straits be determined in a procedurally generated map?

Factorio analysis: data munging

What is the command to reset a PC without deleting any files

Copycat chess is back

The use of multiple foreign keys on same column in SQL Server

Can I make popcorn with any corn?

Extreme, but not acceptable situation and I can't start the work tomorrow morning

Is Social Media Science Fiction?

Schwarzschild Radius of the Universe

Patience, young "Padovan"

Chess with symmetric move-square

Why does this relative pronoun not take the case of the noun it is referring to?

A newer friend of my brother's gave him a load of baseball cards that are supposedly extremely valuable. Is this a scam?

Can you lasso down a wizard who is using the Levitate spell?

How would photo IDs work for shapeshifters?

Does the radius of the Spirit Guardians spell depend on the size of the caster?

How do I create uniquely male characters?

caretEnsemble fails when na.action=na.pass is included

How to track a progress while building model with the caret package?Errors when running Caret package in RError with the train function in caret when using gbm with custom weightscaretEnsmble with different trained datasets with SVM (issue)How to train several models within a loop forHow to pass saved models to caretEnsemblecross validation in r with caret packageR caret gbm classification, predict.gbm(…, type=“response) and the caret package predict(…,type=”prob") of a gbm do not matchError with caret package - classification v regressionHow to fix this error: “Error in auc3_(actual, predicted, ranks) : Not compatible with requested type: [type=character; target=double].”

.everyoneloves__top-leaderboard:empty,.everyoneloves__mid-leaderboard:empty,.everyoneloves__bot-mid-leaderboard:empty height:90px;width:728px;box-sizing:border-box;

1

I want to create a ensemble model using the 'caretList' function in R package 'caretEnsemble'

When I include the code na.action = na.pass in my ensemble code, I get the error

"Error in as.vector(x, "character") : 
 cannot coerce type 'closure' to vector of type 'character'"

As far as I understand this error occurs when you try to use an R function as a object. But my code appears correct and I used the same code to successfully train separate models in R package "caret"

If I exclude na.action=na.pass, I get the error

"Error in na.fail.default(list(y = c(2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L,: missing values in object"

If I run the model on the dataset where I have excluded all the incomplete rows (where any observation has an na), it runs.
My dataset has 19 (all numeric) predictor variables, 30905 rows (observations) with several na values in it (some variables up to 18 000 na's). So I do not want to only model with only complete rows.

My model code is as follows:

set.seed(1269)
 mycontrol<- trainControl(
 method="repeatedcv", number = 5, repeats = 5,
 savePredictions = "final",
 classProbs=TRUE,
 index = createResample(dat$y,25),
 summaryFunction=twoClassSummary
 )

model_list <- caretList(
 y~., data=dat,
 trControl=mycontrol,
 na.action = na.pass,
 methodList=c("ranger","gbm")
 )

I have checked the structure of the data. The predictor variable is a two-class factor and the response variables are all either numeric or integers. I have also used the same data to successfully train independent models in caret, so I doubt it is a problem with the data.

Does anyone have a possible solution for me? Thank you.

Edit: I made use of the function "knnImputation" in the R package "DMwR" to impute NA values before modelling, in stead of including this in the "preprocess" function/phase. Now my model is running. But I am still interested to find out why I had these errors.

r machine-learning na r-caret

share|improve this question

edited Mar 9 at 4:07

jmuhlenkamp

1,508526

asked Mar 1 at 11:21

MiMiMiMi

184128

add a comment | 

1

I want to create a ensemble model using the 'caretList' function in R package 'caretEnsemble'

When I include the code na.action = na.pass in my ensemble code, I get the error

"Error in as.vector(x, "character") : 
 cannot coerce type 'closure' to vector of type 'character'"

As far as I understand this error occurs when you try to use an R function as a object. But my code appears correct and I used the same code to successfully train separate models in R package "caret"

If I exclude na.action=na.pass, I get the error

"Error in na.fail.default(list(y = c(2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L,: missing values in object"

If I run the model on the dataset where I have excluded all the incomplete rows (where any observation has an na), it runs.
My dataset has 19 (all numeric) predictor variables, 30905 rows (observations) with several na values in it (some variables up to 18 000 na's). So I do not want to only model with only complete rows.

My model code is as follows:

set.seed(1269)
 mycontrol<- trainControl(
 method="repeatedcv", number = 5, repeats = 5,
 savePredictions = "final",
 classProbs=TRUE,
 index = createResample(dat$y,25),
 summaryFunction=twoClassSummary
 )

model_list <- caretList(
 y~., data=dat,
 trControl=mycontrol,
 na.action = na.pass,
 methodList=c("ranger","gbm")
 )

I have checked the structure of the data. The predictor variable is a two-class factor and the response variables are all either numeric or integers. I have also used the same data to successfully train independent models in caret, so I doubt it is a problem with the data.

Does anyone have a possible solution for me? Thank you.

Edit: I made use of the function "knnImputation" in the R package "DMwR" to impute NA values before modelling, in stead of including this in the "preprocess" function/phase. Now my model is running. But I am still interested to find out why I had these errors.

r machine-learning na r-caret

share|improve this question

edited Mar 9 at 4:07

jmuhlenkamp

1,508526

asked Mar 1 at 11:21

MiMiMiMi

184128

add a comment | 

I want to create a ensemble model using the 'caretList' function in R package 'caretEnsemble'

When I include the code na.action = na.pass in my ensemble code, I get the error

"Error in as.vector(x, "character") : 
 cannot coerce type 'closure' to vector of type 'character'"

As far as I understand this error occurs when you try to use an R function as a object. But my code appears correct and I used the same code to successfully train separate models in R package "caret"

If I exclude na.action=na.pass, I get the error

"Error in na.fail.default(list(y = c(2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L,: missing values in object"

If I run the model on the dataset where I have excluded all the incomplete rows (where any observation has an na), it runs.
My dataset has 19 (all numeric) predictor variables, 30905 rows (observations) with several na values in it (some variables up to 18 000 na's). So I do not want to only model with only complete rows.

My model code is as follows:

set.seed(1269)
 mycontrol<- trainControl(
 method="repeatedcv", number = 5, repeats = 5,
 savePredictions = "final",
 classProbs=TRUE,
 index = createResample(dat$y,25),
 summaryFunction=twoClassSummary
 )

model_list <- caretList(
 y~., data=dat,
 trControl=mycontrol,
 na.action = na.pass,
 methodList=c("ranger","gbm")
 )

I have checked the structure of the data. The predictor variable is a two-class factor and the response variables are all either numeric or integers. I have also used the same data to successfully train independent models in caret, so I doubt it is a problem with the data.

Does anyone have a possible solution for me? Thank you.

Edit: I made use of the function "knnImputation" in the R package "DMwR" to impute NA values before modelling, in stead of including this in the "preprocess" function/phase. Now my model is running. But I am still interested to find out why I had these errors.

r machine-learning na r-caret

share|improve this question

edited Mar 9 at 4:07

jmuhlenkamp

1,508526

asked Mar 1 at 11:21

MiMiMiMi

184128

"Error in as.vector(x, "character") : 
 cannot coerce type 'closure' to vector of type 'character'"

"Error in na.fail.default(list(y = c(2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L,: missing values in object"

set.seed(1269)
 mycontrol<- trainControl(
 method="repeatedcv", number = 5, repeats = 5,
 savePredictions = "final",
 classProbs=TRUE,
 index = createResample(dat$y,25),
 summaryFunction=twoClassSummary
 )

model_list <- caretList(
 y~., data=dat,
 trControl=mycontrol,
 na.action = na.pass,
 methodList=c("ranger","gbm")
 )

share|improve this question

edited Mar 9 at 4:07

jmuhlenkamp

1,508526

asked Mar 1 at 11:21

MiMiMiMi

184128

share|improve this question

edited Mar 9 at 4:07

jmuhlenkamp

1,508526

asked Mar 1 at 11:21

MiMiMiMi

184128

edited Mar 9 at 4:07

jmuhlenkamp

1,508526

edited Mar 9 at 4:07

jmuhlenkamp

1,508526

asked Mar 1 at 11:21

MiMiMiMi

184128

asked Mar 1 at 11:21

MiMiMiMi

184128