| Title: | Spatio-Temporal Fixation Pattern Analysis |
|---|---|
| Description: | Spatio-temporal Fixation Pattern Analysis (FPA) is a new method of analyzing eye movement data, developed by Mr. Jinlu Cao under the supervision of Prof. Chen Hsuan-Chih at The Chinese University of Hong Kong, and Prof. Wang Suiping at the South China Normal Univeristy. The package "fpa" is a R implementation which makes FPA analysis much easier. There are four major functions in the package: ft2fp(), get_pattern(), plot_pattern(), and lineplot(). The function ft2fp() is the core function, which can complete all the preprocessing within moments. The other three functions are supportive functions which visualize the eye fixation patterns. |
| Authors: | Jinlu Cao |
| Maintainer: | Jinlu Cao <[email protected]> |
| License: | GPL-2 |
| Version: | 1.0 |
| Built: | 2024-11-13 03:12:20 UTC |
| Source: | https://github.com/cran/fpa |
"Spatio-temporal Fixation Pattern Analysis" (FPA) is a new method of analyzing eye movement data, developed by Mr. Jinlu Cao under the supervision of Prof. CHEN Hsuan-Chih at The Chinese Univeristy of Hong Kong, and Prof. Wang Suiping at The South China Normal Univeristy. The method provides a new way to inspect the spato-temporal fixation patterns of eye movements.
| Package: | fpa |
| Type: | Package |
| Version: | 1.0 |
| Date: | 2016-08-13 |
| License: | GPL-2 |
The package "fpa" is a R implementation which makes FPA analysis much easier. There are four major functions in the package: ft2fp(), get_pattern(), plot_pattern(), and lineplot(). The function ft2fp() is the core function, which can complete all the preprocessing within seconds or minutes. The other three functions are supportive functions which visualize the eye fixation patterns.
Jinlu Cao
Maintainer: Jinlu Cao <[email protected]>
ft2fp, get_pattern, plot_pattern, lineplot
data(rawdata) newdata <- ft2fp (rawdata, 4, 3000, 100) pattern <- get_pattern(newdata) plot_pattern(pattern)data(rawdata) newdata <- ft2fp (rawdata, 4, 3000, 100) pattern <- get_pattern(newdata) plot_pattern(pattern)
The ft2fp() function transforms the fixation time (start and end time for each fixation) data to fixation probability data. The function can finish all the preprocessing of using FPA to analyze eye movement data.
ft2fp(data,CriticalRegion,TimeCourse,Interval,norm=TRUE,rm.nr=FALSE,rm.1p=TRUE)ft2fp(data,CriticalRegion,TimeCourse,Interval,norm=TRUE,rm.nr=FALSE,rm.1p=TRUE)
data |
the raw eye movement data provided by user. A data frame which contains variables of "List", "Subject", "Item", "Condition", "Region", "Fix_Start", and "Fix_End". The names and number of variables in your data should be exactly same with above. |
CriticalRegion |
the No. of region in which the researcher is interested. All fixation information before the first-pass on that region will be discarded for each trial. |
TimeCourse |
the time course to be analyzed after the first-pass of critical region. The unit is millisecond. |
Interval |
the time interval (or bin) to show in the time course of interest. The unit is millisecond, and the value should be smaller than the value for TimeCourse. |
norm |
to choose whether to normalize the fixation duration according to each subject's mean duration and general mean duration. If TRUE, the fixation durations are adjusted for each subject's reading rate. The default value if TRUE. |
rm.nr |
to choose whether to exclude the trials with no regression after the first-pass on critical region. The default value is FALSE. |
rm.1p |
to choose whether to exclude the fixations at the first pass (or Gaze duration) on critical region. The default value is TRUE. |
a data frame with the variables of "list", "subject", "condition", "region", "time", "fix_prob" (fixation probability), "y" (number of trials with fixation) and "N" (number of total valid trials).
Jinlu Cao
data(rawdata) newdata <- ft2fp (rawdata, 4, 3000, 100) newdata <- ft2fp (rawdata, 4, 3000, 100, norm=TRUE, rm.nr=TRUE, rm.1p=FALSE)data(rawdata) newdata <- ft2fp (rawdata, 4, 3000, 100) newdata <- ft2fp (rawdata, 4, 3000, 100, norm=TRUE, rm.nr=TRUE, rm.1p=FALSE)
The get_pattern() function aggregates the data so that the general fixation pattern can be shown for each condition. Users should provide the data frame returned in ft2fp() function. Users can use the returned data frame of this function to make plots on the pattern by themselves, or use plot_pattern() and lineplot() functions.
get_pattern(data)get_pattern(data)
data |
is the data frame returned by the ft2fp function. |
a data frame which shows the averaged fixation probabilities for each spatio-temporal unit for each condition.
Jinlu Cao
data(newdata) pattern <- get_pattern(newdata)data(newdata) pattern <- get_pattern(newdata)
The function lineplot() provides quick tools for plotting more detailed fixation probabilities for specific condition(s) and region(s). The function generates 2-dimensional line plots with "Time" as x, and "Fixation Probability" as y.
lineplot(data, Region = "All", Condition = "All")lineplot(data, Region = "All", Condition = "All")
data |
the data frame returned by get_pattern function. |
Region |
the intended region(s) to plot. It can be a string ("All"), a number (e.g., 1), or a vector (e.g., c(1,2)). |
Condition |
the intended condition(s) to plot. It can be a string ("All"), a number (e.g., 1), or a vector (e.g., c(1,2)). |
Jinlu Cao
get_pattern, ft2fp, plot_pattern
data(pattern) lineplot(pattern) lineplot(pattern, Region="All", Condition=1) lineplot(pattern, Condition=c(1,2)) lineplot(pattern, Region=2) lineplot(pattern, Region=c(2,3), Condition=c(3,4,5))data(pattern) lineplot(pattern) lineplot(pattern, Region="All", Condition=1) lineplot(pattern, Condition=c(1,2)) lineplot(pattern, Region=2) lineplot(pattern, Region=c(2,3), Condition=c(3,4,5))
This data set is the fixation probability data generated by the ft2fp() function. The data set retains the information of list, subject, and condition of original fixation time data set. The variable "Time" is generated based on the TimeCourse and TimeInterval arguments defined by users. The variable "N" is the total number of valid trials after the deletion of invalid ones for the corresponding spatio-temporal unit. The variable "y" is the number of trials with fixations on that particular region at that time point. Fixation probability is calculated by dividing y by N. "N" and "y" would be used for further analysis in empirical logistic transformation and lme modeling.
data(newdata)data(newdata)
In the data frame each row represents the fixation probability and other information for one spatio-temporal unit. The data frame has the following columns:
listthe id of the list
subjectthe id of the subject
conditionthe id of the condition
regionthe id of the region
Timethe time after eyes leave critical region
Ntotal number of valid trials
ynumber of trials with fixations
fix_probthe fixation probability
This data set is generated by get_pattern() function, and describes the general fixation pattern for different conditions. The values under time variables are the aggregated fixation probabilities for each spatio-temporal unit. Users may use the data set to make plots of the pattern by themselves or use the dataset as argument of plot_pattern() and lineplot() functions.
data(pattern)data(pattern)
The data set aggregates the fixation probability data, and show the general fixation pattern. The data frame has the following core columns:
conditionthe id of the condition
regionthe id of the region
0averaged fixation probability at time 0
2500averaged fixation probability at time 0
The plot_pattern() function provides a quick tool to plot the fixation pattern for conditions. It generates 3-dimensional data, with x of "Time", y of "Region", and the colors representing the value of fixation probabilities.
plot_pattern(data, Condition = "All")plot_pattern(data, Condition = "All")
data |
the data frame returned by the get_pattern function. |
Condition |
the conditions which the user would like to plot. It can be a string ("All"), a number (e.g., 1), or a vector (e.g., c(1,2)). The default value is "All", meaning all conditions will be plotted. |
Jinlu Cao
data(pattern) plot_pattern(pattern) plot_pattern(pattern, Condition=1) plot_pattern(pattern, Condition=c(1,2))data(pattern) plot_pattern(pattern) plot_pattern(pattern, Condition=1) plot_pattern(pattern, Condition=c(1,2))
This data set is the eye movement data recorded during reading sentences. The sentences consists of 8 regions defined by the researcher. The experiment consists of 2 lists (8 items in each list), 2 subjects, and 4 conditions. Each row contains the information of one fixation. This data frame is a template for the data to be provided by users of fpa-package, and the number and names of variables should be same with this data frame. If the user does not have several lists in his/her experiment, lease use the id 1 for every row.
data(rawdata)data(rawdata)
In the data frame each row represents one fixation of the eyes. Fixations are ordered chronologically within trial. The data frame has the following columns:
Listthe id of the list
Subjectthe id of the subject
Conditionthe id of the condition
Itemthe id of the item
Regionthe id of the region being fixed on
Fix_Startthe start time of the fixation
Fix_Endthe end time of the fixation