CodeToRun.R

# Manage project dependencies ------
# install.packages("renv") # if not already installed, install renv from CRAN
# the following will prompt you to install the various packages used in the study 
renv::activate()
renv::restore() 

# Load packages ------
# load r packages
library(SqlRender)
library(DatabaseConnector)
library(FeatureExtraction) 
library(here)
library(lubridate)
library(stringr)
library(ggplot2)
library(DBI)
library(dbplyr)
library(dplyr)
library(tidyr)
library(kableExtra)
library(RSQLite) 
library(rmarkdown)
library(tableone) 
library(scales)
library(forcats) 
#library(epiR) ## not working
library(RPostgres)
library(cmprsk) #
library(mstate)
library(broom) #
library(rms)
library(glue) #
library(readr)
library(log4r) #
library(survival)
library(flexsurv)
library(tictoc)
library(purrr)
library(CirceR)
library(remotes)
library(CohortGenerator)
library(survminer)


# install.packages("drat")
# drat::addRepo("OHDSI")
# install.packages("FeatureExtraction")

# Set the name/ acronym for your database (to be used in the titles of reports, etc) -----
db.name<-"CPRD_Aurum"

# Set output folder location -----
# the path to a folder where the results from this analysis will be saved
# to set the location within the project with folder called "CPRD", we can use: here("CPRD")
# but this file path could be set to somewhere else
output.folder<-here("Results",db.name)


# database connection details
server     <- Sys.getenv("DB_SERVER_cdm_aurum_202106") # AURUM
server_dbi <- Sys.getenv("DB_SERVER_cdm_aurum_202106_dbi") #AURUM
user       <- Sys.getenv("DB_USER")
password   <- Sys.getenv("DB_PASSWORD")
port       <- Sys.getenv("DB_PORT") 
host       <- Sys.getenv("DB_HOST") 

# Specify OHDSI DatabaseConnector connection details  ------
# set up the createConnectionDetails to connect to the database
# see http://ohdsi.github.io/DatabaseConnector for more details

downloadJdbcDrivers("postgresql", here()) # if you already have this you can omit and change pathToDriver below
connectionDetails <- createConnectionDetails(dbms = "postgresql",
                                             server =server,
                                             user = user,
                                             password = password,
                                             port = port ,
                                             pathToDriver = here())



# Specify DBI connection details -----
# In this study we also use the DBI package to connect to the database
# set up the dbConnect details below (see https://dbi.r-dbi.org/articles/dbi for more details)
# you may need to install another package for this (although RPostgres is included with renv in case you are using postgres)

db <- dbConnect(RPostgres::Postgres(), dbname = server_dbi, port = port, host = host, user = user,
                password = password)


# Set database details -----

# your sql dialect used with the OHDSI SqlRender package
# eg postgresql, redshift etc
# see https://ohdsi.github.io/SqlRender/articles/UsingSqlRender.html for more details
targetDialect <-"postgresql" 

# The name of the schema that contains the OMOP CDM with patient-level data
cdm_database_schema<-"public"

# The name of the schema that contains the vocabularies 
# (often this will be the same as cdm_database_schema)
vocabulary_database_schema<-"public"

# The name of the schema where results tables will be created 
results_database_schema<-"results"

# Tables to be created in your results schema for this analysis will be named using this as the stem 
# Note, any existing tables in your results schema with the same names will be overwritten
cohortTableStem<-"CancerExtrapolation"

# Check database connections -----
# to check whether the OHDSI DatabaseConnector worked, uncomment and run the below three lines
# conn <- connect(connectionDetails)
# querySql(conn,paste0("SELECT COUNT(*) FROM ", cdm_database_schema, ".person"))
# disconnect(conn)

# to check the DBI worked, uncomment and run the below line
#tbl(db, sql(paste0("SELECT * FROM ",cdm_database_schema, ".person"))) %>% tally()

# in both cases, you should have a count of people in the database printed back in the console

# Run the study ------
#source(here("RunStudy.R"))

# after the study is run you should have a zip folder in your output folder to share