Downsample pupil data by integer-factor aggregation

Aggregates consecutive samples within independent sequences. The default mean method is safer than simple decimation because it reduces aliasing and preserves a representative timestamp for each bin.

Usage

downsample_gazepoint_pupil(
  master_df,
  factor = 2,
  pupil_cols = NULL,
  id_col = "USER_ID",
  group_cols = NULL,
  ts_col = "TIME",
  method = c("mean", "first"),
  keep_bin = FALSE
)

Arguments

master_df

A sample-level data frame.

factor

Positive integer downsampling factor.

pupil_cols

Pupil columns to aggregate. When NULL, common pupil columns are detected automatically.

id_col

Participant identifier.

group_cols

Optional additional independent-sequence columns.

ts_col

Optional timestamp column. Its bin value is the finite mean.

method

"mean" aggregation or "first"-sample decimation.

keep_bin

Keep the generated downsample-bin identifier.

Value

A downsampled data frame.

Examples

pupil <- data.frame(
  USER_ID = "P01",
  TIME = seq(0, 0.09, by = 0.01),
  mean_pupil = seq(3, 4, length.out = 10)
)
downsample_gazepoint_pupil(pupil, factor = 2)
#>   USER_ID  TIME mean_pupil n_samples_aggregated downsample_factor
#> 1     P01 0.005   3.055556                    2                 2
#> 2     P01 0.025   3.277778                    2                 2
#> 3     P01 0.045   3.500000                    2                 2
#> 4     P01 0.065   3.722222                    2                 2
#> 5     P01 0.085   3.944444                    2                 2