I'm getting the following error message

"exception while shutting down RxClientPipe: Write error: expected 8 got _4294967295"

when executing the sample code:

# Multi-class logistic regression
testObs <- rnorm(nrow(iris)) > 0
testIris <- iris[testObs,]
trainIris <- iris[!testObs,]
multiLogit <- rxLogisticRegression( formula = Species~Sepal.Length + Sepal.Width + Petal.Length + Petal.Width, type = "multiClass", data = trainIris)
# Score the model score
MultiDF <- rxPredict(multiLogit, data = testIris, extraVarsToWrite = "Species")
# Print the first rows of the data frame with scores
head(scoreMultiDF)
# Look at confusion matrix
table(scoreMultiDF$Species, scoreMultiDF$PredictedLabel)
# Look at the observations with incorrect predictions
badPrediction = scoreMultiDF$Species != scoreMultiDF$PredictedLabel
scoreMultiDF[badPrediction,]

What would be the base url for the request to find the Access Token?
The body seems to contain the username & password. Is there any other detail that needs to be passed in the body?

I am using a Data Science Virtual Machine that has "version.string R version 3.4.3 (2017-11-30)" installed which according to mrsdeploy documentation is what is required to install the package. Is the documentation out of date? I have created an enterprise configuration using your instructions and have no way to log into it.

I receive

install.packages("mrsdeploy")
Warning in install.packages :
package ‘mrsdeploy’ is not available (for R version 3.4.3)

Hi!

0

I would like to consume a webservice in R Studio, made in XML. The created web service has three components: 1. XXX, 2. YYY and 3. ZZZ

Within each component is the data that is needed.

Currently I am using the httr library, and I can do, let's say, the connection with the data well since I get the status 200.

However, I am not getting the correct data since I am not sure how to correctly specify the correct component (for example, XXX).

The code:

suppressWarnings(library(httr))
suppressWarnings(library(XML))

url <-"https://thelink.com"

username <- "FFF"
password <- "EEE"

MH_1 <- GET(url, authenticate(username,password, type = "basic"))

MH_1
content(MH_1)
rawToChar(MH_1$content)

I don't know how to specify the component (for example XXX) of the WS with the GET() function... could someone please help me. Also, is there another way to consume a WS in R that I could use in this case? ... under the above specifications ...

Thanks!

When will an R Client version be released that can work with more recent versions of R Open. Specifically when will the R Client be compatible with a 3.5.x version of R Open.

We need an update to this product, as it has not been updated since February 2018.

The documentation page Server Compatibility would benefit from links to the downloads to the old versions of Microsoft R Client.

hello everyone!
Maybe (and probably) I'm wrong, but I think that the component "coefLabelStyle" produces an error when you select coefLabelStyle="R".
I was testing the results of each level between rxGlm and glm, and I figured out that when you have no observations in one of the levels then rxGlm mix the other factor values.

I'd like translate this series articles to Chinese. Would you create a localization branch?

olapR links are broken:
https://github.com/MicrosoftDocs/machine-learning-server-docs/blob/live/sql/machine-learning/r/ref-r-olapr

where can I find olapR package code?

Hi

Whilst running step3_training_evaluation.R code from:
r-server-hospital-length-of-stay

Cannot find anything on MicrosoftRML package. Is this a typo in the code?

See error message

Error in rxTlcBridge(formula = lengthofstay ~ rcount + gender + dialysisrenalendstage + :
MicrosoftRML package must be installed.
Calls: source ... eval -> eval -> rxRemoteCall -> do.call -> rxTlcBridge
Execution halted
Error in rxCompleteClusterJob(hpcServerJob, consoleOutput, autoCleanup) :
No results available - final job state: failed

Train the Boosted Trees model.
boosted_model <- rxFastTrees(formula = formula,
data = LoS_Train,
type = c("regression"),
numTrees = 40,
learningRate = 0.2,
splitFraction = 5/24,
featureFraction = 1,
minSplit = 10)
XXX ( process 1 ) has started run at 2020-03-16 15:45:44.41 ======
Error in rxTlcBridge(formula = lengthofstay ~ rcount + gender + dialysisrenalendstage + :
MicrosoftRML package must be installed.
Calls: source ... eval -> eval -> rxRemoteCall -> do.call -> rxTlcBridge
Execution halted
Error in rxCompleteClusterJob(hpcServerJob, consoleOutput, autoCleanup) :
No results available - final job state: failed

It would be immensely beneficial if this document provided more information on how a Python service is deployed, beyond the single example given at https://github.com/Microsoft/ML-Server-Python-Samples/blob/master/operationalize/Quickstart_Publish_Python_Web_Service.ipynb.
As far as I can see, the following questions are not answered in the documentation.

• If deploying a Python model referencing code in several Python files, where should the referenced Python files be stored so that the deployed model can find them? (For example, say I'm deploying a Python function to do basic arithmetic. The deployed model calls a function "add" from the file add-util.py. Where should add-util.py be stored?)
• Where is a deployed model stored in the machine learning server?
• When defining a Python service to deploy (as in the link above), what is the difference between the code_fn function and models function?
• There is documentation on offline R package installs, but nothing for offline Python packages installs (from a whl file, probably). Can this be added?

If there is existing documentation on any of these questions, please redirect me.