How do I update the plane or add a new dataset?

[mrchrisadams]

I'm interested in adding a dataset for calculating the carbon footprint of servers, as it's not clear to me which one I would use when I do a search on carbon kit here.

https://www.carbonkit.net/search?q=computer

What does this process look like?

My current understanding is that we'd need ALL of these to make a new one

• itemdef.csv - an abstract description of the components of the dataset (i.e. table columns) and methodology (inputs, outputs, return values)
• data.csv - a table of data, usually describing specific scenarios relating to an emissions-producing activity
• default.js - a javascript algorithm which describes the computation of some output (usually emissions) based on certain inputs and dataset data
• return_values.csv - description of any calculation output values. In some cases this may be absent and only a single output is provided based upon the final evaluated statement in the default.js algorithm.
• documentation.creole - text file contain a written description of the dataset/methodology and which appears in discover.amee.com
• meta.yml - metadata for the dataset/methodology include source information, annotations and tags
• changelog.yml - a description of changes made

But it's not clear to me where in the hierarchy of datasets this would go.

### Looking for a good example to start with

In terms of datasets, I've looked at this one before, as it seems quite small and easy to understand:

https://github.com/OpenAMEE/datasets/blob/master/business/agriculture/livestock/entericfermentation/documentation.creole

And I've shared some links to my understanding of it here:

https://slides.com/mrchrisadams/an-npm-for-environmental-data#/4/8

Once you've written a new dataset.

Once I've written one, I'm not sure how I'd test it though.

I reckon you might be able to mock out the the call here to simulate running it in the platform:

var ch4GWP = parseFloat(dataFinder.getDataItemValue('planet/greenhousegases/gwp','gas=CH4','GWP'));

What would be needed to make a PR for a new, test dataset?

_Aside: In fact, if you I knew what variables were expected to be in the context of running the default.js file, i reckon you might be able to run it by itself.