get_bins()

Introduction

The use of digital systems for behavior measurement often provides detailed insights into various dimensions of observed phenomena. However, the resolution of these measurements may vary depending on the hardware and software employed. When analyzing data from time-related experiments, it is frequently essential to organize data points into “bins” of a specified time unit. This allows for flexibility in adjusting the resolution at which the data will be analyzed.

In the case of the proprietary data output format from the ‘MED-PC’ software suite, time is often recorded with a time.event structure. Depending on how the software is set up, time will often be recorded with a tenth of a second resolution, having data points like 436.12, 534,3 756.1 and so on. Consequently, it becomes necessary to group the time information into more manageable data sets.

Here we introduce the get_bins() function which allows to “binarize” a numeric vector with a given resolution.

The get_bins() function takes the following parameters:

• x numeric the numeric data points vector to be binarized.
• x_min the minimum value for the bin distribution.
• x_max the maximum value of the vector “x” to binarize.
• res the numeric value for the resolution (e.g., 1s).

It is important to have in mind the time resolution you’re working with; for example, if your data is on a tenth of a second resolution and you want to binarize with a 1s resolution, res should be 10.

Example

First let’s create a random set of numeric values from 0 to 600 to simulate a 1 minute trial with a tenth of seconds resolution.

set.seed(420)
sim_time <- runif(35, 0, 600) # Generates 35 random data points between 0 and 600.
sim_time <- round(sim_time) # Rounds the resulting numbers to work with integers.
sim_time <- sort(sim_time) # Arrange the data in ascending order.

sim_time

##  [1]   1  99 105 108 118 133 141 174 182 188 194 202 227 229 232 262 285 289 304
## [20] 319 363 385 393 422 425 430 435 442 446 496 504 526 528 550 582

Now we apply the get_bins() to binarize the distribution with a 2 seconds resolution.

binarized_data <- get_bins(sim_time, 0, 600, 20) # Note how the resolution value is 20

binarized_data

##  [1]  20 100 120 120 120 140 160 180 200 200 200 220 240 240 240 280 300 300 320
## [20] 320 380 400 400 440 440 440 440 460 460 500 520 540 540 560 600

Now the original time data is represented as their respective 2s time bin.