Read image data from a collection of csvs from TimeLapse

This function reads in image data from a directory of csv files, or a character vector of csv file paths. It attempts to identify the appropriate TimeLapse template to use based on the file names, but a specific template can also be provided. The function does some basic standardization and cleaning of the data, and returns a data.frame with an "image_data" class and an attribute "template" with the name of the template used to read the data.

Usage

read_image_data(path, pattern, recursive = FALSE, template = NULL, ...)

Arguments

path: path to directory of image files, a single .csv file, or a character vector of .csv files.
pattern: an optional regular expression. Only file names which match the regular expression will read. Default FALSE.
recursive: should files found within subfolders of path also be read?
template: path to a ".tdb" TimeLapse Template file. Optional; if not provided, the function will attempt to identify the appropriate internal template based on the file names in path.
...: arguments passed on to readr::read_csv()

Value

a data.frame of Timelapse image data from the files found in path. The data.frame will have an "image_data" class, and an attribute "template" with the name of the template used to read the data.

Details

In addition to reading in the data, this function also:

does basic standardization of trigger mode values,
creates numeric snow depth columns,
checks for the presence of a total_count_episode column, adding one if missing,
standardizes column names to snake_case.
If the data has separate Date and Time columns instead of a combined DateTime column, these will be reconciled into a single DateTime column.

For wallow data, this also removes motion-activated static images, and only keeps the motion-activated video records and the time-lapse static image records.