OHDSI Home | Forums | Wiki | Github

PatienLevelPrediction Installation check response codes

patientprediction

(Dramacloak) #1

I’m running into some response codes that indicate issues with different aspects of PatientLevelPrediction.

require(devtools)
require(curl)
require(rJava)

#devtools::install_github(“OHDSI/OhdsiRTools”, force=TRUE)
require(OhdsiRTools)
#devtools::install(“F:/GitHub OHDSI/SqlRender-master”, quick=TRUE, force=TRUE)
require(SqlRender)
#devtools::install(“F:/GitHub OHDSI/DatabaseConnector-master”, quick=TRUE, force=TRUE)
require(DatabaseConnector)
#devtools::install(“F:/GitHub OHDSI/Achilles-master”, quick=TRUE, force=TRUE)
require(Achilles)
#devtools::install(“F:/GitHub OHDSI/DatabaseConnectorJars-master”,quick=TRUE, force=TRUE)
require(DatabaseConnectorJars)
#devtools::install(“F:/GitHub OHDSI/FeatureExtraction-master”,quick=TRUE, force=TRUE)
require(FeatureExtraction)
#devtools::install(“F:/GitHub OHDSI/OhdsiSharing-master”,quick=TRUE, force=TRUE)
require(OhdsiSharing)
#devtools::install(“F:/GitHub OHDSI/BigKnn-master”,quick=TRUE, force=TRUE)
require(BigKnn)
#devtools::install(“F:/GitHub OHDSI/PatientLevelPrediction-master”,quick=TRUE, force=TRUE)
require(PatientLevelPrediction)
#devtools::install_github(“rstudio/keras”)
require(keras)
#install_keras()

connectionDetails <- createConnectionDetails(
dbms=“postgresql”,
server=“localhost/OHDSI”,
user=“postgres”,
password=’’,
port=“5432”)

options(fftempdir = “F:/Github OHDSI/fftemp”)
PatientLevelPrediction::checkPlpInstallation(connectionDetails = connectionDetails, python = F)

Checking database connectivity
Connecting using PostgreSQL driver

  • Done

Checking R population
Generating covariates
Generating cohorts
Generating outcomes
Generating exclusion

  • Done

Checking Models

  • Ok
  • Done

Checking support for large data objects

  • Done

PatientLevelPrediction installation check completed…

Response code: 88319
Issue with createStudyPopulation()
Issue with lasso logistic regression
Issue with knn
Issue with gradient boosting machine
NULL

PatientLevelPrediction::checkPlpInstallation(connectionDetails = connectionDetails, python = T)
Checking database connectivity
Connecting using PostgreSQL driver

  • Done

Checking R population
Generating covariates
Generating cohorts
Generating outcomes
Generating exclusion

  • Done

Checking Models

  • Ok
  • Done

Checking support for large data objects

  • Done

PatientLevelPrediction installation check completed…

Response code: 247357937827
Issue with createStudyPopulation()
Issue with lasso logistic regression
Issue with random forest
Issue with mlp
Issue with ada boost
Issue with decison tree
Issue with naive bayes
Issue with knn
Issue with gradient boosting machine
NULL

system(“python --version”)
Python 3.6.7 :: Anaconda, Inc.

I’m running in Windows 7 on R 3.5.1 using a Postgresql database. Any ideas what’s wrong or is this output normal?


(Dramacloak) #2

This seems to manifest when I try to run Lasso regression as follows:
covariateSettings <- createCovariateSettings(useDemographicsGender = TRUE
)

plpData <- getPlpData(connectionDetails = connectionDetails,
cdmDatabaseSchema = “cdm”,
cohortDatabaseSchema = “results”,
cohortTable = ‘cohort’,
cohortId = 5,
covariateSettings = covariateSettings,
outcomeDatabaseSchema = “results”,
outcomeTable = “cohort”,
outcomeIds = 2,
sampleSize = 500)

plpDatb <- getPlpData(connectionDetails = connectionDetails,
cdmDatabaseSchema = “cdm”,
cohortDatabaseSchema = “results”,
cohortTable = ‘cohort’,
cohortId = 6,
covariateSettings = covariateSettings,
outcomeDatabaseSchema = “results”,
outcomeTable = “cohort”,
outcomeIds = 2,
sampleSize = 1200)

#population <- createStudyPopulation(plpData = plpData, outcomeId = 3, riskWindowStart = 30, includeAllOutcomes = TRUE, verbosity = “DEBUG”)

lrmodel <- setLassoLogisticRegression(seed = 50)
lrResults <- runPlp(plpDatb, plpData, modelSettings = lrModel, testSplit =‘person’,testFraction = 0.25,nfold=2,splitSeed = 1234)
Error: evaluation nested too deeply: infinite recursion / options(expressions=)?


(Dramacloak) #3

These errors have disappeared after I installed the Cohort Method and RiskStratifiedEstimation packages and the test runs without critical error (though there are quite a few “is deprecated” messages).


(Dramacloak) #4

It was working for a bit, but now the issue has resurfaced. Strange. Any idea how to get this working again?


(Jreps) #5

I’ve not seen that issue before, it seems odd that the createStudyPopulation() is also failing. Can you try installing using a different function:
devtools::install_github(“OHDSI/PatientLevelPrediction”)

and if that still fails try the branch I’m working on:
devtools::install_github(“OHDSI/PatientLevelPrediction”,ref = “efficient” )

Also, if you post issues here: https://github.com/OHDSI/PatientLevelPrediction/issues I’ll get alerted, on the forum I might miss things.

I just searched the error and it seems to be a problem when using recursion. I don’t think I’ve got any recursion in the PLP functions but maybe a dependency is doing it. It might get solved if you run:
options(expressions=1000)


t