Schemes

In this section we present the schemes and formats that the different files of this project must have.

CSV files

Comma-separated value (CSV) files can be used with the csv2json.py script in order to create an experiment file from a data file. However, to create a correct experiment file, the data file must follow some formatting rules:

  • One row corresponds to one experimental run

  • One column correspond to one factor

  • In term of columns ordering, the columns representing the factors must come before the columns representing the response variable(s)

  • The first row of the data must contain the variable names, with the unit of the variable in parenthesis at the end (e.g. “Applied pressure (Pa)”)

  • The columns must be separated by semi-colon

  • For numerical values with decimal numbers, use a comma ‘,’ as the decimal separator

Only the first three formatting rules are mandatory. The other ones are simply the default options but can be changed when using csv2json.py.

Experiment files

Experiment files are .yaml files that follow a specific structure.

Validation

We can use the same script we use in github actions to locally validate experiment files. This script can be found in utils/validation/scripts/validation.py in this repository.

Calling

python utils/validation/scripts/validation.py --file experiments/test.yaml

we will be validating as experiment the contents of test.yaml.

Calling the script without –file will validate all the yaml files contained in experiments/ folder that were changed in the current working tree and the main branch tree. We use this functionality in the github action validation pipeline.

Validation schema

show here the schema against which we validate our experiment Work in progress…

Keywords validation

To validate that all the keywords present in the experiment file are described in the glossary, we can use another validation script. This script can be found in utils/validation/scripts/keywords_validation.py in this repository.

Calling

python utils/validation/scripts/keywords_validation.py

will simply notify you if any of the keywords are unexplained.