Expand description
Irregular functional data operations.
This module provides data structures and algorithms for functional data where observations have different evaluation points (irregular/sparse sampling).
§Storage Format
Uses a CSR-like (Compressed Sparse Row) format for efficient storage:
offsets[i]..offsets[i+1]gives the slice indices for observation iargvalsandvaluesstore all data contiguously
This format is memory-efficient and enables parallel processing of observations.
§Example
For 3 curves with varying numbers of observation points:
- Curve 0: 5 points
- Curve 1: 3 points
- Curve 2: 7 points
The offsets would be: [0, 5, 8, 15]
Structs§
- Irreg
Fdata - Compressed storage for irregular functional data.
Functions§
- cov_
irreg - Estimate covariance at a grid of points using local linear smoothing.
- integrate_
irreg - Compute integral of each curve using trapezoidal rule.
- integrate_
irreg_ struct - Compute integral for IrregFdata struct.
- mean_
irreg - Estimate mean function at specified target points using kernel smoothing.
- metric_
lp_ irreg - Compute pairwise Lp distances for irregular functional data.
- norm_
lp_ irreg - Compute Lp norm for each curve in irregular functional data.
- to_
regular_ grid - Convert irregular data to regular grid via linear interpolation.