Carlos Ramos Carreño
Carlos Ramos Carreño
Consider adding auxiliary functions for common plotting utilities (and maybe other parts too) to hide unnecessary OOP from users not accustomed to it.
Plotting bands for a percent of the deepest data could be useful in several contexts. We could add functionality to do just that.
Currently, the number of basis used for a particular dataset must be explicitly selected. It would be nice to add functionality to choose this number according to some criterion.
Some methods in functional data analysis can deal with multivariate functional data (several functions per observation, not necessarily in the same domain) and mixed data (multivariate measures and functional data...
**Is your feature request related to a problem? Please describe.** Currently, scikit-fda does not have any type for irregularly sampled data. A fully-fledged irregular type will require weeks or months...
I open this issue to list missing examples that should be in the documentation: - [x] A good FDataBasis tutorial, presenting each basis and comparing between them. - [x] UEA...
The mRMR implementation should be tested against the original algorithm (using the same estimator for mutual information). The references to the original papers should be added to the docs.
Implement scoring methods for functional response (to be used for example in validation and hyperparameter search).
**Describe the bug** Currently when the domain range of a basis is not specified at creation time, the method `to_basis` adjust this to the domain range of the functions. This...
**Is your feature request related to a problem? Please describe.** Lasagna plots are another representation of functional data (in the form of curves). It is specially useful to visualize the...