# Summary
**See single_source/README.md the latest approach for creating a single source file from a CSV. Other
files are kept to compare the latest approach against if interested**

This directory contains resources for parsing the single source spreadsheet to produce a JSON file containing measure eligibility and performance information.



This file is then used during the enrichment of the measures data to include those attributes.
# Running
## Containerised
### Prerequisites
The only prerequisite to run the containerised process is that you have [docker](https://www.docker.com/) and [docker-compose](https://docs.docker.com/compose/) installed.
### Instructions
Run the script from the container:
```bash
docker-compose run data-analysis python /home/jovyan/work/scripts/single_source_to_json.py
```
The docker-compose file in this directory mounts the correct directories for this operation using the default paths. Other paths can be specified with the following arguments:

* `--single_source_filepath`
    * Path to the spreadsheet containing the claims-related data, e.g. `/claims-related/data/2017_Claims_IndivMeasures_SingleSourceVersion2_11082017_andrea_edit`
    * Note : If clean up is needed, after the creation of a CSV from XLSX, place the original in the archive directory. Clean up can be performed (for empty rows, junk characters etc) with clean_csv.py or manually.
* `--output_filepath`
    * Path to write output to, e.g. `/claims-related/data/qpp-single-source.json`
### Local development
When making changes to the single source parsing scripts, use the command `make test` to run the test suite. Doing this ensures that any changes do not break existing functionality, resulting in an unusable or inaccurate JSON output.
## Native
The script for processing the spreadsheet can be run outside of the container using the instructions above if your system has [numpy](http://www.numpy.org/) and [pandas](https://pandas.pydata.org/) installed.