fdaPDE-cpp
fdaPDE-cpp copied to clipboard
fdaPDE
A C++ library for physics-informed spatial and functional data analysis over complex domains.
In applications it might be necessary to consider responses and/or covariates defined on different spatial grids and/or different areal partitionings, e.g. we have a covariate defined on a regular grid,...
The implementation of method `P()` for `SpaceTimeParabolicBase` in fdaPDE/models/space_time_parabolic_base.h is missing. This prevents, e.g., the definition of the GCV index for a model using such regularization.
In fPCA, currently, either we set a fixed level of smoothing for all the components, or we are forced to perform a GCV or KCV selection component by component. The...
As for 956ba75, `fpirls_init()` initializes the status of SQR-PDE for the first FPIRLS iteration using the normalized SR-PDE functional ```math J(f, \boldsymbol{\beta}) = \frac{1}{n} \sum_{i=1}^n (y_i - \boldsymbol{x}_i^\top \boldsymbol{\beta} -...
The computation of `J` in [fdaPDE/models/regression/fpirls.h](fdaPDE/models/regression/fpirls.h) is not correct for space-time problems. Specifically, the computation of the penalty term is always set to $\lambda_{\mathcal{D}} \boldsymbol{g}^\top R_0 \boldsymbol{g}$, which is ok...
Several optimizations can be turned on if we keep a run-time set of flags related to a model instance. The idea would be to declare a public member struct `runtime`...
Pull request for MQSRPDE (Multiple Quantile Spatial Regression with PDE regularization) model to fdaPDE-cpp develop branch
…r parameters have been intrduced. A base class (infernece_base.h) and derived classes for specific methods have been created: wald.h, speckman.h, esf.h, pesf.h. Also an auxilary class for some inference methods...
