updated docs

README.md

-# GOSe 6mo imputation for CENTER-TBI 
+# GOSe imputation for CENTER-TBI 
 
 This repository contains the code for imputing the 6 months GOSe using
 a multi-state model in CENTER-TBI.
-The workflow is completely automated using snakemake and a singularity
-container with the required dependencies is awailable at [zenodo.org](https://zenodo.org/record/2641161) 
-(DOI: 10.5281/zenodo.2641161).
+Execution requires 
 
-Singularity >= 2.6.0 is required to run the code in a container, alternatively
-the dependencies can be installed manually (check the `container-recipe` file).
+1. access to the CENTER-TBI Neurobot database at https://center-tbi.incf.org/ or
+a manually downloaded data file according to the specification given below
+2. a recent version of [R](https://www.r-project.org/) (4+)
+3. (optionally) the [RStudio IDE](https://rstudio.com/products/rstudio/download/)
+4. (optionally) the version control system [git](https://git-scm.com/) to 
+easily download this repository to your local file system.
 
-To get started, download the code and `cd` in the repository folder
+To get started, download the code and change into the repository folder.
+Using git on a Unix system (Linux) this can be done via
 ```bash
 git clone https://git.center-tbi.eu/kunzmann/gose-6mo-imputation-production.git
 cd gose-6mo-imputation-production
 ```
-Next, download the pre-build container image from zenodo.org via the 
-provided script
-```bash
-./scripts/download-container
+Open this folder in Rstudio. This should activate the prepared 
+R environment automatically and locall;y install any R package dependencies. 
+See the documentation for [renv](https://rstudio.github.io/renv/articles/renv.html)
+in case of problems.
+The required R environment can also be restored manually by invoking
 ```
-This will download the `container.sif` file (~1Gb) in the current working 
-directory.
-An automated download script via the Neurobot-API is provided but requires
-that the Neurobot username and API token are available as environment variables
-```bash
-export NEUROBOT_USR=[your-neurobot-username]
-export NEUROBOT_API=[your-neurobot-api-token]
+install.packages("renv")
+renv::restore(clean = TRUE, prompt = FALSE)
 ```
-The required data set (for version 1.2 on the Neurobot staging server) can 
-then automatically be downloaded via
-```bash
-singularity exec container.sif snakemake download
+in an R console with working directory set to `[my-working-direcrtory]/gose-6mo-imputation-production` 
+where `[my-working-direcrtory]` is the location on yur file system that you 
+placed the repository in.
+
+To execute the markdown file the recorded GOSe values are required 
+in a .csv file named `tbl_gose.csv` in the same folder, i.e. `[my-working-direcrtory]/gose-6mo-imputation-production/tbl_gose.csv`.
+
+The .csv file must contain the Neurobot variables 
 ```
-Alternatively, a .csv file with the correct column names / ordering can 
-directly be places in the `data` subfolder (see `scripts/check-inputs` for 
-detailed specification).
-The imputation can then be started by invoking
-```bash
-singularity exec container.sif snakemake impute
+subjectId
+Outcomes.DerivedCompositeGOSE
+Outcomes.DerivedCompositeGOSEDaysPostInjury
+Subject.Age
+Subject.PatientType
+Subject.DeathDate
 ```
-This produces a .html and a .csv file with imputed values in the subfolder 
-`output`.
+in exactly this order. 
+At the time of writing this report the newest staging data set could be
+accessed via the saved link `https://neurobot-stage.incf.org/_5f476d41eedb867816a10376`.
+
+Finally, open the project in Rstudio, open the `imputation-report.Rmd` in
+Rstudio and [`knit' it](https://rmarkdown.rstudio.com/).
+This produces a .html with a basic report and a `tbl_gose_imputed.csv` 
+file with imputed values.
 

imputation-report.Rmd

@@ -212,7 +212,7 @@ fit <- msm::msm(
   censor.states = 2:7, # cannot be dead since these were filtered previously 
   control   = list(
       fnscale = 12000,
-      maxit   = 10,# 10^4,
+      maxit   = 10^4,
       trace   = 2
   )
 )