Test Environment for FASTGenomics Docker Images

Docker files and descriptions on how to run interactive and batch analysis. Useful as a reference for developing and testing new analysis.

Usage

Clone the test-data into a newly created data folder within this repository
```
git clone https://github.com/FASTGenomics/test-data data
```
Alternatively, supply your own test data - but remember to use the proper FASTGenomics directory structure (e.g. ./data/dataset_0001/my_dataset.loom) and provide a dataset_info.json. Update the docker-compose file with the path to the data set (should point to a directory containing the dataset_0001 folder in the example above).
Clone your favorite analysis (analysis.ipynb) into the analysis folder, e.g.,
```
git clone https://github.com/FASTGenomics/analysis_empty_scanpy analysis
```
You can use any other path on your local machine, just remember to update the corresponding docker-compose.yml path. All the data in the analysis folder will be available in your analysis. Our available analyses can be found here
If you use an existing image from dockerhub specify it in the docker-compose.yml.
Alternatively, if you want to build and test an image - clone it to an image folder and uncomment the build: ... option in docker-compose.yml.
The latest versions of our standard images (fastgenomics/jupyter-scanpy and fastgenomics/jupyter-seurat) can be found on dockerhub.
If you want to develop your own images make sure to always use the latest fastgenomics/jupyter-base version as a strating point.

If you are developing an image you will need to re-build it every time you make some changes to the images source code. To do this run

docker-compose build

To run an analysis in a batch-mode (non-interactive) simply run

docker-compose up batch

This will generate an output file analysis/analysis.html.

To run an analysis in an interactive mode use

docker-compose up interactive