| getHdata {Hmisc} | R Documentation |
This function downloads and makes ready to use datasets from the main
web site for the Hmisc and Design libraries. For R, the datasets were
stored in compressed save format and getHdata makes them
available by running load() after download. For S-Plus, the
datasets were stored in data.dump format and are made available
by running data.restore() after import. The dataset is run through the
cleanup.import function to reduce multiple inheritance problems
for SV4 (S-Plus 5 or later). Calling getHdata with no
file argument provides a character vector of names of available
datasets that are currently on the web site. For R, R's default browser
can optionally be launched to view html files that were already
prepared using the Hmisc command html(contents( )) or to view
.txt or .html data description files when available.
getHdata(file, what = c("data", "contents", "description", "all"),
where="http://biostat.mc.vanderbilt.edu/twiki/pub/Main/DataSets")
file |
an unquoted name of a dataset on the web site,
e.g. prostate. Omit file to obtain a list of
available datasets. |
what |
specify what="contents" to browse the contents
(metadata) for the dataset rather than fetching the data
themselves. Specify what="description" to browse a data
description file if available. Specify what="all" to
retrieve the data and see the metadata and description. |
where |
URL containing the data and metadata files |
For S-Plus, Hmisc defines a function download.file that is used
by getHdata. This is a stripped-down version of the R
download.file function that uses the system wget
executable for fetching files from the Internet. For Unix and Linux
systems, wget will be pre-installed usually. For windows
S-Plus systems, get wget from
ftp://sunsite.dk/projects/wget/windows. Once you unzip the file
from there, move wget.exe to the same Windows directory that
contains ftp.exe.
getHdata() without a file argument returns a character
vector of dataset base names. When a dataset is downloaded, the data
frame is placed in search position one and is not returned as value of
getHdata.
Frank Harrell
download.file, cleanup.import,
data.restore, load
## Not run:
getHdata() # download list of available datasets
getHdata(prostate) # downloads, load( ) or data.restore( )
# runs cleanup.import for S-Plus 6
getHdata(valung, "contents") # open browser (options(browser="whatever"))
# after downloading valung.html
# (result of html(contents()))
getHdata(support, "all") # download and open one browser window
datadensity(support)
attach(support) # make individual variables available
getHdata(plasma, "all") # download and open two browser windows
# (description file is available for plasma)
## End(Not run)