| test.psych {psych} | R Documentation |
Test to make sure the psych functions run on basic test data sets
test.psych(first=1,last=5,short=TRUE,all=FALSE,fapc=FALSE)
first |
first=1: start with dataset first |
last |
last=5: test for datasets until last |
short |
short=TRUE - don't return any analyses |
all |
To get around a failure on certain Solaris 32 bit systems, all=FALSE is the default |
fapc |
if fapc=TRUE, then do a whole series of tests of factor and principal component extraction and rotations. |
When modifying the psych package, it is useful to make sure that adding some code does not break something else. The test.psych function tests the major functions on various standard data sets. It also shows off a number of the capabilities of the psych package.
Uses 5 standard data sets:
USArrests Violent Crime Rates by US State (4 variables)
attitude The Chatterjee-Price Attitude Data
Harman23.cor$cov Harman Example 2.3 8 physical measurements
Harman74.cor$cov Harman Example 7.4 24 mental measurements
ability.cov$cov 8 Ability and Intelligence Tests
It also uses the bfi and ability data sets from psych
out |
if short=FALSE, then list of the output from all functions tested |
Warning messages will be thrown by fa.parallel and sometimes by fa for random datasets.
Although test.psych may be used as a quick demo of the various functions in the psych packge, in general, it is better to try the specific functions themselves. The main purpose of test.psych is to make sure functions throw error messages or correct for weird conditions.
The datasets tested are part of the standard R data sets and represent some of the basic problems encountered.
When version 1.1.10 was released, it caused errors when compiling and testing on some Solaris 32 bit systems. The all option was added to avoid this problem (since I can't replicate the problem on Macs or PCs). all=TRUE adds one more test, for a non-positive definite matrix.
William Revelle
#test <- test.psych() #not run #test.psych(all=TRUE) # f3 <- fa(bfi[1:15],3,n.iter=5) # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="Varimax") # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="varimax") # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="bifactor") # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="varimin") # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="bentlerT") # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="geominT") # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="equamax") # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="Promax") # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="cluster") # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="biquartimin") # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="equamax") # f3 <- fa(bfi[1:15],3,n.iter=5,rotate="Promax") # # fpoly <- fa(bfi[1:10],2,n.iter=5,cor="poly") # f1 <- fa(ability,n.iter=4) # f1p <- fa(ability,n.iter=4,cor="tet")