Readme

.env -sample

@@ -0,0 +1,4 @@
+USERNAME = "my_username"
+PASSWORD = "azerty123"
+AUTH_TOKEN ="TOKEN"
+ROOT_FOLER = "/home/"

.gitignore

@@ -1,10 +1,6 @@
 *.csv
 *.odt
 *.sif
-.Rproj.user
-R_baseline/
-TODO.md
-config.py
 *.zip
 *.mp4
-References/
+.env

README.md

@@ -1,40 +1,41 @@
 # Template-Python-Singularity
-## A template to build executable Singularity images in Python projects
+## A template to build executable Singularity images for Python projects
 
 ### Objective
 
-This repo contains a template to create a Python app that will run in an executable Singularity image. A Singularity containers can typically be used in data science projects to provide transparency and support of reproducibility.
+This repo contains a template to create a Python app that will run in an executable Singularity container. A Singularity container can typically be used in data science projects to provide transparency and support of reproducibility.
 
 ### Content
 
-* `singularity/`: singularity container
-  * singularity_test
-* `utils/` : utility scripts (to make submissions, etc)
-
-* `src/`: models
+* `utils/` : utility scripts. The code in this directory is not included in the singularity container.
+  * `build_image.sh` : a script to build a container image based on `singularity/recipe.txt`. The corresponding Singularity image is executable, and contains the scripts in `src/`. By default, the execution of the container image runs `src/model.py` (edit `recipe.txt` to change this behavior).
+  * `test_container.sh` : a script to test if the execution of the container image works properly.
+* `singularity/`: contains the singularity image built with `build_image.sh`
+  * `singularity_test` : a directory where a singularity image is copied and tested.
+* `src/`: the code that is included in the container.
 * `data/` : input data
 * `output/` : where code output is stored
 
 ### Instruction to build the singularity container from the recipe.txt
 
-To create an executable singularity container (which contains the code in /src), run :
+To create an executable singularity container (which contains the code in `/src`) from `singularity/recipe.txt`, run :
 
 ```bash
 ./utils/build_container.sh
 ```
 
-This will create a container container.sif in /singularity.
-To test the executable singularity container, run :
+This will create a container image `/singularity/container.sif`.
+To test the executable singularity container image, run :
 
 ```bash
 ./utils/test_container.sh
 ```
 
-This will output the results in singularity/test_singularity/output.
+This will output the results of `src/model.py` in `singularity/test_singularity/output`.
 
-### Instructions to run the model
+### Instructions to execute the image
 
-The singularity container container.sif is executable, you can reproduce the results by running:
+The singularity container image container.sif is executable, you can execute it by running:
 
 ```bash
 singularity run ./singularity/container.sif <Argument for input path>
@@ -43,11 +44,10 @@ singularity run ./singularity/container.sif <Argument for input path>
 Note: `<Argument for input path> ` is optional, default is ./
 
 
-Note that container.sif should be in the same folder as /training and /test.
-The prediction scores on test dataset are saved in `./output/scores.csv`.
 ### Instructions to run a custom model in the same environment
 
-To run a model locally on synthetic data in the container environement, activate the singularity container in Terminal using:
+By default, the script called model.py is executed upon execution of the image.
+To run a specific Python script locally on synthetic data in the container environment, activate the singularity container in Terminal using:
 
 ```bash
 singularity shell ./singularity/container.sif
@@ -57,22 +57,15 @@ This will allow you to have an environment (`Singularity>`) to run the model wit
 In order to run the model, run the following command :
 
 ```bash
-Singularity> python3 src/model.py <Argument for input path>
-```
-
-For example, on my computer, when my cwd is the DreamHF folder, I run :
-
-```bash
-Singularity> python3 src/model.py ./
+Singularity> python3 src/script.py <Argument for input path>
 ```
 
 This command can also be used to run the singularity without accessing the singularity shell:
 
 ```bash
-singularity exec container.sif python3 src/model.py <Argument for input path>
+singularity exec container.sif python3 src/script.py <Argument for input path>
 ```
 
-
 ### License
 
 This software is distributed under the GNU GENERAL PUBLIC LICENSE. Please refer to the file LICENCE.txt included for details.