dilation_speed_outliers_to_na: Set dilation speed outliers to missing

In WillemSleegers/eyepatch:

Description

dilation_speed_outliers_to_na calculates the dilation speed based on timestamps and pupil size measurements and then sets outliers to missing (NA).

Usage

dilation_speed_outliers_to_na(pupil, time, constant = 10)

Arguments

pupil	A numeric vector of pupil size measurements.
time	A vector containing the timestamps associated with the pupil size measurements.
constant	A numeric value specifying the threshold for outlier removal. The default is 10.

Details

The exact method of calculating the dilation speed is taken from Kret & Sjak-Shie (2018).

Examples

# Load the "dplyr", "tidyr", and "ggplot2" packages:
library(dplyr)
library(tidyr)
library(ggplot2)

# Example 1: Artificial data
# Create some artificial data:
data <- tibble(
  time = 0:5,
  pupil_left = c(3.11, 3.13, 3.17, 3.7, 3.16, 3.12),
  pupil_right = c(2.92, 2.95, 2.99, 3.5, 2.97, 2.95)
)
data

# Remove outliers:
mutate(data,
  pupil_left = dilation_speed_outliers_to_na(pupil_left, time,
    constant = 10),
  pupil_right = dilation_speed_outliers_to_na(pupil_right, time,
    constant = 10)
)

Example 2: Realistic data
blink

# Remove outliers:
blink <- mutate(blink,
  pupil_left_new = dilation_speed_outliers_to_na(pupil_left, timestamp,
    constant = 10),
  pupil_right_new = dilation_speed_outliers_to_na(pupil_right, timestamp,
    constant = 10)
)

# Restructure the data and plot the results to compare the pupil measurements
# before the outlier removal and after the outlier removal:
blink %>%
  rename(
    pupil_left_old = pupil_left,
    pupil_right_old = pupil_right
  ) %>%
  pivot_longer(
    cols = starts_with("pupil"),
    names_to = c("pupil", "status"),
    names_pattern = "(left|right)_(old|new)",
    values_to = "pupil_size"
 ) %>%
 ggplot(aes(x = timestamp, y = pupil_size, color = status)) +
   geom_point() +
   facet_wrap(~ pupil)

WillemSleegers/eyepatch documentation built on Aug. 2, 2021, 8:39 a.m.