guess.experiment.filenames: Guessing Experiment File Organization

In haberdashPI/wright.lab.util: A set of utilitiy functions used in our lab.

Description

Guess the organization of files for a given experiment. This guessworks is based on naming conventions used throughout old programs run in the lab. Each file is a single block of an expeirment, and indicates the subject id, date, and task for the block.

Usage

guess.experiment.filenames(dir, file.pattern = old.file.pattern, default.month.codes = month.codes, subject.month.codes = list(), label.post.test = NULL,verified.ids = NULL)

Arguments

dir	the directory containing all files form a single expeirment
file.pattern	The naming convention used by files. There are two predefined patterns, old.file.pattern and new.file.pattern. Old files types start with letters and new files start with numbers. You can make your own file patterns. Refer to file.patterns.R for more information.
default.month.codes	the default month code used to order files month codes can be NULL, in which case files are sorted by their date modified. The month codes translate strings into an unambiguous format so that sorting by the date in the file name will correctly order files. There are two sets of pre-made month codes: month.codes and late.month.codes month.codes should work for most files but if there are late months that have been ambiguously coded, this will fail. For subject with ambiguous late months you should use set subject specific month code. See the file.patterns.R file for more information on making your own month codes.
subject.month.codes	provdies subject specific codes for converting months to an unambiguous format. Each item in the list has the name of a subject, and a value with month codes used to convert dates into sortable format for the given subejct.
label.post.test	If non-NULL this will cause the last day for each participant to be set to its value. Thus all post tests will have the same value, and can be easilly selected. This is useful to set when some participants practice a different number of days than others.
verified.ids	Beause dates are ambiguously coded in many cases, it is important to verify that they are actually in the correct order. For each subject not added to this list, a warning will be printed asking you to verify that the order is correct. Once you have verified a given subject, add their id to this list. BUT MAKE SURE THE ID IS UNIQUE. Otherwise if you add new subjects there will be no warning message to alert you to the possibility that dates are our of order. If a subject's days are not ordered correctly you have several options, listed below from easiest to hardest. 1. You can specify a different set of month codes to use for that subject. Often you just need to give them the late.month.codes instead of month.codes. 2. You can use the date modified of each file to order days. You can do this by specifying a file pattern with 'day' set to NULL. 3. You can also design your own month codes (see file.patterns.R) for more information. 4. You can re-order the days for this subject in the returned data table from this function.

Value

Returns a data set that contains a row for each file. The columns indicate the subject, task and day for that file. You can use find.file.deltas to determine the threshold for each data file in this data set.

Author(s)

David Little

Examples

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

haberdashPI/wright.lab.util documentation built on Nov. 4, 2019, 1:25 p.m.