mosaic (version 0.4-1)

fetchData: A Web and Library Data-Loading Facility

Description

fetchData provides a means for students and others to locate and load data sets provided by instructors. Data can be pre-loaded for off-line sessions, can be positioned on identified web sites, or loaded from packages. fetchData also will load local .csv files using file.choose().

Usage

fetchData(name = NULL, add.to.path = FALSE,
    drop.from.path = FALSE, add.to.library = FALSE,
    directory = NULL, data = NULL, verbose = TRUE)

Arguments

name

a character string naming a data set. This will often end in .csv for reading in a data set. When used in conjunction with TRUE values for the following arguments, it can also name a web directory (always ending in

add.to.path

If TRUE, indicates that the web search path is to printed out, or, if name is specified, the name should be a web directory (ending in /), which should be pre-pended to the search path.

drop.from.path

If TRUE, wipes out the web search path, or, if name is specified, removes that web directory from the search path.

add.to.library

If TRUE, indicates that a data set is to be pre-loaded into the cached library. This allows, for instance, users to pre-load on-line data to be used when they are off-line.

verbose

a logical indicating whether additional status messages (e.g., indicating where the dataset was located) should be printed.

data

The data frame to be put in the cached library if add.to.library=TRUE.

directory

The name of a web directory to be searched but not added to the search path.

Value

• a data frame.

Details

There are two major purposes for this function. One is to provide a consistent interface to reading data: a file name is given and a data frame is returned, which can be assigned to an object as the user desires. This differs from the behavior of data, which doesn't return a value but instead creates an object without explicit assignment.

The other purpose is to allow instructors or other group leaders to post data on web sites that can be searched as naturally as if the data were on the users' own machines. For instance, an instructor might want to post a new data set just before class, enabling her students to access it in class.

To support this, fetchData allows new web sites to be added to the web search path. Typically, the command to add a site would be in a script file that is provided to the student that could be run automatically at start up or sourced over the web. That is, an instructor might create a script file stored on a website and, using a web page, provide students with the text of the command to source it.

Examples

kids <- fetchData("KidsFeet.csv")
carbon <- fetchData("CO2")
fetchData(add.to.path=TRUE)
fetchData(add.to.path=TRUE, name="http://www.macalester.edu/~kaplan/ISM/datasets/")
fetchData(drop.from.path=TRUE, name="http://www.macalester.edu/~kaplan/ISM/datasets/")
fetchData(drop.from.path=TRUE)
fetchData(add.to.library=TRUE, name="mydata.csv", data=data.frame(x=c(1,2,3), y=c(7,1,4)))