Skip to contents

Getting started with {improveR}

improveR.Rmd

Overview

improveR is the R interface to Scinteco’s improve platform - a secure, versioned repository for pharmaceutical R&D workflows. It provides seamless integration between R and the improve platform, enabling 21 CFR Part 11 compliant workflow management for pharmaceutical modeling, biostatistical analysis, and data science.

Key capabilities include:

  • Workflow Management: Create and execute multi-step pharmaceutical modeling workflows
  • Version Control: Full versioning of data, code, and analyses with complete audit trails
  • Resource Management: Load and save analytical resources with metadata and lineage tracking
  • Collaborative Research: Team-based review system for pharmaceutical research

Installation

Install improveR from the Scinteco repository:

# Add the improve repository
options(repos = c(getOption("repos"),
  improve = "https://r.scinteco.com/4.4.0/"))

# Install the package
install.packages("improveR")

# Optional: Install CLI tools
install.packages("improveRcontributions")

For detailed installation instructions, see vignette("install-improveR").

Setup and Connection

Configure your environment variables to connect to the improve repository:

# Add to your .Renviron file:
# Sys.setenv(IMPROVER_REPO_URL = "https://<url>:<port>/repository")
# Sys.setenv(IMPROVER_STEP = "<valid-entityId>")

# Load the package
library(improveR)

# Connect to the repository
improveConnect()

# Enable editing
setEditable(TRUE)

# Verify connection
whoami()

Core Concepts

Resources

The improve platform organizes everything as resources - versioned entities with unique identifiers:

  • Folders: Organize resources hierarchically
  • Files: Store data, scripts, and outputs with automatic versioning
  • Links: Create stable references to specific versions of resources
  • Steps: Executable units of analytical work within workflows
# Load a resource by path
resource <- loadResource("/Projects/my-analysis/data")

# Load by entity ID
resource <- loadResource("EN-12345")

# List folder contents
children <- loadChildResources(resource)

# Create a new file
myData <- data.frame(x = 1:10, y = rnorm(10))
write.csv(myData, "data.csv", row.names = FALSE)

dataFile <- createFile(
  targetIdent = resource$entityId,
  fileName = "data.csv",
  localPath = "data.csv",
  comment = "My analysis data"
)

Workflows

Workflows chain together analytical steps organized in analysis trees:

# Create an analysis tree
analysisTree <- createAnalysisTree(
  targetIdent = "/Projects/my-analysis",
  treeName = "EDA_Workflow",
  comment = "Exploratory data analysis"
)

# Create a step environment
stepEnv <- createStepTemplateEnv(treeIdent = analysisTree)

# Configure the step
stepEnv$setStepDescription("Load and prepare data")
stepEnv$setStepRationale("Initial data loading and validation")

# Add input files
stepEnv$addStepRemoteFile(ident = dataFile$entityId, variableName = "input-data")

# Set execution configuration
stepEnv$setStepRunserverLabel("runserver")
stepEnv$setStepToolLabel("R_4.2")
stepEnv$setStepToolInstance("rbatch")

# Create the step
resultStep <- stepEnv$realise()
step <- resultStep$getStepResource()

Use getStep() to explore workflow relationships:

# Get a step with its full context
stepEnv <- getStep("/Projects/my-analysis/EDA_Workflow/Step 1")

# Load and explore child steps
stepEnv$children$load()
ls(stepEnv$children)

# Load and explore lineage (upstream dependencies)
stepEnv$lineage$load(treeDepth = -1)
workflow <- stepEnv$workflow$df()

# Load usage (downstream consumers)
stepEnv$usage$load(treeDepth = -1)

# Check step state
stepEnv$getStepState()

# Get step inventory
inventory <- stepEnv$getStepInventory()

Next Steps

Now that you understand the basics, explore these topics in more detail:

Getting Help 

# Function documentation
?improveConnect
?createFile
?getStep

# List all improveR functions
ls("package:improveR")

# Disconnect when done
improveDisconnect()

For support, visit the Scinteco website.