| sasxport.get {Hmisc} | R Documentation |
Uses the read.xport and lookup.xport functions in the
foreign library to import SAS datasets. SAS date, time, and
date/time variables are converted respectively to Date,
POSIX, or chron objects in R,
variable names are converted to lower case, SAS labels are associated
with variables, and (by default) integer-valued variables are converted
from storage mode double to integer. If the user ran
PROC FORMAT CNTLOUT= in SAS and included the resulting dataset in
the SAS version 5 transport file, variables having customized formats
that do not include any ranges (i.e., variables having standard
PROC FORMAT; VALUE label formats) will have their format labels looked
up, and these variables are converted to S factors.
For those users having access to SAS, method='csv' is preferred
when importing several SAS datasets.
Run SAS macro exportlib.sas available from
http://biostat.mc.vanderbilt.edu/twiki/pub/Main/Hmisc/exportlib.sas
to convert all SAS datasets in a SAS data library (from any engine
supported by your system) into CSV files. If any customized
formats are used, it is assumed that the PROC FORMAT CNTLOUT=
dataset is in the data library as a regular SAS dataset, as above.
SASdsLabels reads a file containing PROC CONTENTS
printed output to parse dataset labels, assuming that PROC
CONTENTS was run on an entire library.
sasxport.get(file, force.single = TRUE,
method=c('read.xport','dataload','csv'), formats=NULL, allow=NULL,
out=NULL, keep=NULL, drop=NULL, as.is=0.5, FUN=NULL)
sasdsLabels(file)
file |
name of a file containing the SAS transport file.
file may be a URL beginning with http://. For
sasdsLabels, file is the name of a file containing a
PROC CONTENTS output listing. For method='csv',
file is the name of the directory containing all the CSV
files created by running the exportlib SAS macro.
|
force.single |
set to FALSE to keep integer-valued
variables not exceeding 2^31-1 in value from being converted to
integer storage mode |
method |
set to "dataload" if you have the dataload
executable installed and want to use it instead of
read.xport. This seems to correct some errors in which
rarely some factor variables are always missing when read by
read.xport when in fact they have some non-missing values. |
formats |
a data frame or list (like that created by
read.xport) containing PROC FORMAT
output, if such output is not stored in the main transport file. |
allow |
a vector of characters allowed by R that should not be converted to periods in variable names. By default, underscores in variable names are converted to periods as with R before version 1.9. |
out |
a character string specifying a directory in which to write
separate R save files (.rda files) for each regular
dataset. Each file and the data frame inside it is named with the
SAS dataset name translated to lower case and with underscores
changed to periods. The default NULL value of out
results in a data frame or a list of data frames being returned.
When out is given, sasxport.get returns only metadata (see
below), invisibly.
out only works with methods='csv'. out should
not have a trailing slash. |
keep |
a vector of names of SAS datasets to process (original SAS
upper case names). Must include PROC FORMAT dataset if it
exists, and if the kept datasets use any of its value label formats. |
drop |
a vector of names of SAS datasets to ignore (original SAS upper case names) |
as.is |
SAS character variables are converted to S factor
objects if as.is=FALSE or if as.is is a number between
0 and 1 inclusive and the number of unique values of the variable is
less than the number of observations (n) times as.is.
The default if as.is is .5, so character variables are
converted to factors only if they have fewer than n/2 unique
values. The primary purpose of this is to keep unique
identification variables as character values in the data frame
instead of using more space to store both the integer factor codes
and the factor labels.
|
FUN |
an optional function that will be run on each data frame
created, when method='csv' and out are specified. The
result of all the FUN calls is made into a list corresponding
to the SAS datasets that are read. This list is the FUN
attribute of the result returned by sasxport.get.
|
See contents.list for a way to print the
directory of SAS datasets when more than one was imported.
If there is more than one dataset in the transport file other than the
PROC FORMAT file, the result is a list of data frames
containing all the non-PROC FORMAT datasets. Otherwise the
result is the single data frame. There is an exception if out
is specified; that causes separate R save files to be written
and the returned value to be a list corresponding to the SAS datasets,
with key PROC CONTENTS information in a data frame making up
each part of the list.
sasdsLabels returns a named
vector of dataset labels, with names equal to the dataset names.
Frank E Harrell Jr
read.xport,label,sas.get,
Dates,DateTimeClasses,chron,
lookup.xport,contents,describe
## Not run:
# SAS code to generate test dataset:
# libname y SASV5XPT "test2.xpt";
#
# PROC FORMAT; VALUE race 1=green 2=blue 3=purple; RUN;
# PROC FORMAT CNTLOUT=format;RUN; * Name, e.g. 'format', unimportant;
# data test;
# LENGTH race 3 age 4;
# age=30; label age="Age at Beginning of Study";
# race=2;
# d1='3mar2002'd ;
# dt1='3mar2002 9:31:02'dt;
# t1='11:13:45't;
# output;
#
# age=31;
# race=4;
# d1='3jun2002'd ;
# dt1='3jun2002 9:42:07'dt;
# t1='11:14:13't;
# output;
# format d1 mmddyy10. dt1 datetime. t1 time. race race.;
# run;
# data z; LENGTH x3 3 x4 4 x5 5 x6 6 x7 7 x8 8;
# DO i=1 TO 100;
# x3=ranuni(3);
# x4=ranuni(5);
# x5=ranuni(7);
# x6=ranuni(9);
# x7=ranuni(11);
# x8=ranuni(13);
# output;
# END;
# DROP i;
# RUN;
# PROC MEANS; RUN;
# PROC COPY IN=work OUT=y;SELECT test format z;RUN; *Creates test2.xpt;
w <- sasxport.get('test2.xpt')
# To use an existing copy of test2.xpt available on the web:
w <- sasxport.get('http://hesweb1.med.virginia.edu/biostat/s/data/sas/test2.xpt')
describe(w$test) # see labels, format names for dataset test
# Note: if only one dataset (other than format) had been exported,
# just do describe(w) as sasxport.get would not create a list for that
lapply(w, describe)# see descriptive stats for both datasets
contents(w$test) # another way to see variable attributes
lapply(w, contents)# show contents of both datasets
options(digits=7) # compare the following matrix with PROC MEANS output
t(sapply(w$z, function(x)
c(Mean=mean(x),SD=sqrt(var(x)),Min=min(x),Max=max(x))))
## End(Not run)