SED-ML Aggregation function (New terms for count, stddev, variance)

[jonrkarr]

• count: number of non-zero elements
• mean
• median
• standard deviation
• variance
• length

[matthiaskoenig]

• standard error

would be great in addition

[matthiaskoenig]

Can we add the remaining functions under https://bioportal.bioontology.org/ontologies/KISAO/?p=classes&conceptid=http%3A%2F%2Fwww.biomodels.net%2Fkisao%2FKISAO%23KISAO_0000824? Is there a documentation how to update terms? I could add these myself to get up to speed how to do this.

• [x] maximum (KISAO_0000828): The maximum of a set of values, ignoring NaN entries.
• [x] minimum (KISAO_0000829 The minimum of a set of values, ignoring NaN entries.
• [x] mean (KISAO_0000825) The mean of a set of values, ignoring NaN entries.
• [x] standard deviation (KISAO_0000826): The standard deviation of a set of values, ignoring NaN entries.
• [x] standard error (KISAO_0000827): The standard error of a set of values, ignoring NaN entries.
• [ ] maximum (not ignoring NaN): The maximum of a set of values.
• [ ] minimum (not ignoring NaN): The minimum of a set of values.
• [ ] mean (not ignoring NaN) The mean of a set of values.
• [ ] standard deviation (not ignoring NaN): The standard deviation of a set of values.
• [ ] standard error (not ignoring NaN): The standard error of a set of values.
• [ ] sum: The sum of array elements over a given axis, ignoring Nan entries
• [ ] sum (not ignoring NaN):
• [ ] product: The product of array elements over a given axis, ignoring NaN entries.
• [ ] product (not ignoring NaN): The product of array elements over a given axis, ignoring NaN entries.
• [ ] cumsum: Return the cumulative sum of the elements along a given axis, ignoring NaN entries.
• [ ] cumsum (not ignoring NaN): Return the cumulative sum of the elements along a given axis.
• [ ] cumprod: Return the cumulative product of elements along a given axis, ignoring NaN entries.
• [ ] cumprod (not ignoring NaN): Return the cumulative product of elements along a given axis.
• [ ] count: number of non-zero elements, ignoring NaN entries
• [ ] count (not ignoring NaN): number of non-zero elements, ignoring NaN entries
• [ ] length: number of elements, ignoring NaN entries
• [ ] length (not ignoring NaN): number of elements, ignoring NaN entries
• [ ] median
• [ ] median (not ignoring NaN)
• [ ] variance
• [ ] variance (not ignoring NaN)

It would be amazing if we could change the labels so that we have:

• mean (ignoring NaN)
• mean

which is in line with how this normally works. I.e. the ignoring NaN functions are special versions of the normal functions.

[jonrkarr]

I put a brief guide to editing the ontology here: https://github.com/SED-ML/KiSAO/blob/dev/CONTRIBUTING.md#modifying-the-ontology. Protege is a good editing tool. This makes it easy to see the relationships as you add terms. The existing terms are also a good reference.

Please also add a note about your changes to CHANGELOG.md.

The GitHub action will automatically run the few tests.

I can take of releasing your changes as v2.22.

[jonrkarr]

I agree with labels such as mean and mean (ignoring NaN). This is similar to the names of NumPy functions (mean, nanmean) and the default configuration of MATLAB (mean(..., nanflag)). The "default" should be to include NaN such the result of a mean over a NaN is NaN.

[matthiaskoenig]

The following terms have been added/updated. Pull request shortly:

• [x] maximum ignoring NaN (KISAO_0000828)
• [x] minimum ignoring NaN (KISAO_0000829
• [x] mean ignoring NaN (KISAO_0000825)
• [x] standard deviation ignoring NaN (KISAO_0000826)
• [x] standard error ignoring NaN (KISAO_0000827)
• [x] maximum (KISAO_0000830)
• [x] minimum (KISAO_0000840)
• [x] mean (KISAO_0000841)
• [x] standard deviation (KISAO_0000842)
• [x] standard error (KISAO_0000843)
• [x] sum ignoring NaN (KISAO_0000844)
• [x] sum (KISAO_0000845)
• [x] product ignoring NaN (KISAO_0000846)
• [x] product (KISAO_0000847)
• [x] cumulative sum (KISAO_0000848)
• [x] cumulative sum ignoring NaN (KISAO_0000849)
• [x] cumulative product ignoring NaN (KISAO_0000850)
• [x] cumulative product (KISAO_0000851)
• [x] count ignoring NaN (KISAO_0000852): number of non-zero elements, ignoring NaN entries
• [x] count (KISAO_0000853): number of non-zero elements, ignoring NaN entries
• [x] length ignoring NaN (KISAO_0000854): number of elements, ignoring NaN entries
• [x] length (KISAO_0000855): number of elements, ignoring NaN entries
• [x] median ignoring NaN (KISAO_0000856)
• [x] median (KISAO_0000857)
• [x] variance ignoring NaN (KISAO_0000858)
• [x] variance (KISAO_0000859)