prep_testing_data: Align test data and apply fPCA using elastic method applied...

In veesa: Pipeline for Explainable Machine Learning with Functional Data

Align test data and apply fPCA using elastic method applied to training data

Description

Applies steps 2 and 3 of the VEESA pipeline (alignment and elastic fPCA (jfpca, hfpca, or vfpca)) to the testing data based on the training data prepared using "prep_training_data".

Usage

prep_testing_data(f, time, train_prep, optim_method = "DP")

Arguments

f	Matrix (size M x N) of test data with N functions and M samples.
time	Vector of size M describing the sample points
train_prep	Object returned from applying "prep_training_data" to training data.
optim_method	Method used for optimization when computing the Karcher mean. "DP", "DPo", and "RBFGS".

Value

List containing (varies slightly based on fpca method used):

• time: vector of times when functions are observed (length of M)

• f0: original test data functions - matrix (M x N) of N functions with M samples

• fn: aligned test data functions - similar structure to f0

• q0: original test data SRSFs - similar structure to f0

• qn: aligned test data SRSFs - similar structure to f0

• mqn: training data SRSF mean (test data functions are aligned to this function)

• gam: test data warping functions - similar structure to f0

• coef: test data principal component coefficients

• psi: test data warping function SRVFs - similar structure to f0 (jfpca and hfpca only)

• nu: test data shooting functions - similar structure to f0 (jfpca and hfpca only)

• g: test data combination of aligned and shooting functions (jfpca only)

Examples

# Load packages
library(dplyr)
library(tidyr)

# Select a subset of functions from shifted peaks data
sub_ids <-
  shifted_peaks$data |>
  select(data, group, id) |>
  distinct() |>
  group_by(data, group) |>
  slice(1:4) |>
  ungroup()

# Create a smaller version of shifted data
shifted_peaks_sub <-
  shifted_peaks$data |>
  filter(id %in% sub_ids$id)

# Extract times
shifted_peaks_times = unique(shifted_peaks_sub$t)

# Convert training data to matrix
shifted_peaks_train_matrix <-
  shifted_peaks_sub |>
  filter(data == "Training") |>
  select(-t) |>
  mutate(index = paste0("t", index)) |>
  pivot_wider(names_from = index, values_from = y) |>
  select(-data, -id, -group) |>
  as.matrix() |>
  t()

# Obtain veesa pipeline training data
veesa_train <-
  prep_training_data(
    f = shifted_peaks_train_matrix,
    time = shifted_peaks_times,
    fpca_method = "jfpca"
  )

# Convert testing data to matrix
shifted_peaks_test_matrix <-
  shifted_peaks_sub |>
  filter(data == "Testing") |>
  select(-t) |>
  mutate(index = paste0("t", index)) |>
  pivot_wider(names_from = index, values_from = y) |>
  select(-data, -id, -group) |>
  as.matrix() |>
  t()

# Obtain veesa pipeline testing data
veesa_test <- prep_testing_data(
  f = shifted_peaks_test_matrix,
  time = shifted_peaks_times,
  train_prep = veesa_train,
  optim_method = "DP"
 )

veesa documentation built on April 3, 2025, 6:03 p.m.