After this PR, ionics_fits supports fitting datasets with arbitrary x-axis and y-axis dimensions!
Todo:
[x] update docs to highlight fitting of arbitrary dims
[x] add note that Model2D is really a way of aggregating along x. We've special-cased this to taking a pair of 1D models as input but it would be simple to work with input models of arbitrary dimensions, there just hasn't been a use-case for that yet so we kept it simple
[x] find and replace for the word "channel" with "axis" to complete the renaming of y channels to axes
[x] rename containers to transformations
[x] add test with 2D repeated model to check stacking works
[x] fix validator windowing for multi-dimensional data
[x] finish updating RepeatedModel for multi-dimensional x-axis data (currently broken in terms of parameter scaling
[x] fix axis name conventions (y v z) in multi-d models
[x] Model2D fix scale funcs
[x] Model2D parameter estimator should be able to deal with parameter estimator failing for a single point
After this PR,
ionics_fitssupports fitting datasets with arbitrary x-axis and y-axis dimensions!Todo:
Model2Dis really a way of aggregating along x. We've special-cased this to taking a pair of 1D models as input but it would be simple to work with input models of arbitrary dimensions, there just hasn't been a use-case for that yet so we kept it simpleRepeatedModelfor multi-dimensional x-axis data (currently broken in terms of parameter scaling