tests/testthat/test_smoothSpline.R

In santaR: Short Asynchronous Time-Series Analysis

context('stats::smooth.spline')


if ((as.numeric(R.Version()$major) >= 4) && (as.numeric(R.Version()$minor) >= 2.0) &&
    (.Platform$OS.type == "windows")) {

  ## Current fit of all 8 ind, for all 22 variables, from df=2 to 7
  result_allSmoothSplineFit  <- vector("list", 6)
  # iterate df
  for (df in 2:7) {
    tmp_varFit  <- vector("list", 22)
    # iterate variables
    for (var in 1:22) {
      indName           <- unique(acuteInflammation$meta$ind)
      tmp_indFit        <- vector("list", 8)
      names(tmp_indFit) <- indName
      # interate individuals
      for (ind in 1:8) {
        tmp_x             <- acuteInflammation$meta[acuteInflammation$meta$ind == indName[ind], 'time']
        tmp_y             <- acuteInflammation$data[acuteInflammation$meta$ind == indName[ind], var]
        tmp_indFit[[ind]] <- stats::smooth.spline(x=tmp_x, y=tmp_y, df=df)
      }
      names(tmp_varFit)[var]  <- paste('var', var, sep='_')
      tmp_varFit[var]         <- list(tmp_indFit)
    }
    names(result_allSmoothSplineFit)[df-1]  <- paste('df', df, sep='_') # df is 2:7, list idx to fill are 1:7
    result_allSmoothSplineFit[df-1]         <- list(tmp_varFit)
  }

  ## Expected data
  path_expected_data      <- system.file("testdata/all_smoothSpline_fit.RData", package = "santaR")
  load(path_expected_data) # expected_allSmoothSplineFit

    
  test_that('stats::smooth.spline() fit is consistent', {
    # current vs reference smooth.spline fit
    expect_equal(result_allSmoothSplineFit, expected_allSmoothSplineFit, tolerance=1e-6)
  })
}