R-Executor

Other Language Version: English, 中文.

Author: AFAN (WeChat: afan-life)

R-Executor is a Docker-based R language execution environment that integrates the R runtime and a Python Flask service. Through HTTP requests, users can send R code to the r-executor container for execution and retrieve the results. This project aims to provide a cross-platform R language invocation solution, simplifying the remote execution of R code and result retrieval.

Features

• Flask Integration: Embeds a Flask service into the R Docker environment, creating a remotely callable R execution service.
• Data Type Conversion: Utilizes rpy2 to convert R data types into standard Python data types, enabling HTTP-based result parsing.

Directory Structure

- app.py               # Flask service entry point
- config.py            # Global configuration file
- Dockerfile           # Docker image build file
- requirements.txt     # Python dependency list for building the image
- r_packages.txt       # R dependency list for building the image
- compose.yaml         # Docker Compose configuration file
- test/                # Test cases
  - test0_local.py     # Test R code execution inside the container
  - test1_basic.py     # Test client requests to retrieve R data types
  - test2_lrm.py       # Test client requests to call the lrm function from the rms package for logistic regression
  - test3_plot.py      # Test client requests to generate and save a nomogram plot
- train.csv            # Dataset used for test cases
- nom.png              # Output of test3_plot.py (nomogram plot)

Usage

Get the Image

Pulling the Image from Docker Hub

The r-executor image is available on Docker Hub. You can pull it directly using the following command:

docker pull afanlife/r-executor

Image URL：https://hub.docker.com/r/afanlife/r-executor

The r-executor image comes pre-installed with Python 3.13.2 and R 4.4.2, along with the default Python and R dependencies listed in requirements.txt and r_packages.txt. If the current version meets your needs, you can directly pull the image from Docker Hub.

Note: The app.py in r-executor is executed within the container via volume mounting, allowing direct modifications without rebuilding the image.

Rebuilding the Image

If the default version does not meet your requirements, you can rebuild the image as needed:

• To change the Python or R version, modify the Dockerfile.
• To adjust Python or R dependencies, modify requirements.txt and r_packages.txt.

docker build -t r-executor:1.0 .

Commands for exporting and loading the image:

docker save -o r-executor.tar r-executor
docker load -i r-executor.tar

Starting the Container

Map all files in the current directory to the /app directory inside the container and start the service.

Manual Start

docker run -it -p 5000:5000 -v .:/app afanlife/r-executor /bin/bash
python app.py

Using Docker Compose

docker compose up

Test Cases

Starting app.py

root@9f2585313bab:/app# Python app.py
3.5.17
 * Serving Flask app 'app'
 * Debug mode: on
WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
 * Running on all addresses (0.0.0.0)
 * Running on http://127.0.0.1:5000
 * Running on http://172.17.0.2:5000
Press CTRL+C to quit
 * Restarting with stat

Running test1_basic.py (Testing Basic Data Types)

{
  "data": {
    "dataframe": {
      "a": [
        1.0,
        2.0
      ],
      "b": [
        "x",
        "y"
      ]
    },
    "matrix": [
      1,
      2,
      3,
      4
    ],
    "vector": [
      1.0,
      2.0,
      3.0
    ]
  },
  "message": "R code run success",
  "status": "success"
}

Running test2_lrm.py (Testing Logistic Regression Model)

{
  "data": {
    "accuracy": [
      0.991
    ],
    "model_summary": [
      "Logistic Regression Model",
      "",
      "lrm(formula = Heart_Disease ~ ., data = r_df)",
      "",
      "                       Model Likelihood     Discrimination    Rank Discrim.    ",
      "                             Ratio Test            Indexes          Indexes    ",
      "Obs          1000    LR chi2     238.07     R2       0.859    C       0.997    ",
      " 0            968    d.f.             4    R2(4,1000)0.209    Dxy     0.994    ",
      " 1             32    Pr(> chi2) <0.0001    R2(4,92.9)0.919    gamma   0.994    ",
      "max |deriv| 2e-05                           Brier    0.007    tau-a   0.062    ",
      "",
      "               Coef      S.E.    Wald Z Pr(>|Z|)",
      "Intercept      -109.6707 24.8000 -4.42  <0.0001 ",
      "Age               0.4953  0.1116  4.44  <0.0001 ",
      "Blood_Sugar       0.5558  0.1228  4.52  <0.0001 ",
      "Blood_Pressure    0.0135  0.0213  0.63  0.5270  ",
      "Weight            0.0647  0.0307  2.11  0.0352  ",
      ""
    ]
  },
  "message": "R code run success",
  "status": "success"
}

Running test3_plot.py (Testing Plot Generation and Saving)

References

• r-base Image Build