Thanks to Kit Baum as always, the module cpcorr on SSC has been updated. This module was first posted on SSC (and indeed last revised) in 2001, so long ago that the original post has disappeared into the oblivion of history.
The module requires Stata 10, but so as not to disqualify those who could have made use of the previous version, the original programs remain, but with suffix *6 to indicate that the programs work in Stata 6 through 9.
The changes include:
1. Modernised presentation of the help.
2. Cross-references in the help to other related commands.
3. Saved results.
The module includes an eponymous command cpcorr to give selected rectangular blocks of correlations that need not be entire correlation matrices. An easy example is a vector restricted to the response variable and possible predictors. A more esoteric example is a matrix with rows the original variables and columns their principal components: I find such correlations easier to think about, often, than loadings.
Also included, but not illustrated here, is a cpspear command for Spearman correlation.
Somewhat deeper than these commands, which are just convenience commands, is the minor machinery of how to do it. Let's illustrate with Kendall's tau_a to compensate for not going all the way and providing a command for that measure. In essence, you set up a matrix of the right dimensions and then loop over row and column variables to populate it with results.
The module requires Stata 10, but so as not to disqualify those who could have made use of the previous version, the original programs remain, but with suffix *6 to indicate that the programs work in Stata 6 through 9.
The changes include:
1. Modernised presentation of the help.
2. Cross-references in the help to other related commands.
3. Saved results.
The module includes an eponymous command cpcorr to give selected rectangular blocks of correlations that need not be entire correlation matrices. An easy example is a vector restricted to the response variable and possible predictors. A more esoteric example is a matrix with rows the original variables and columns their principal components: I find such correlations easier to think about, often, than loadings.
Also included, but not illustrated here, is a cpspear command for Spearman correlation.
Code:
. clear
. sysuse auto
(1978 Automobile Data)
. cpcorr headroom trunk weight length displacement \ mpg
(obs=74)
mpg
headroom -0.4138
trunk -0.5816
weight -0.8072
length -0.7958
displacement -0.7056
. cpcorr headroom trunk weight length displacement \ mpg, format(%4.3f)
(obs=74)
mpg
headroom -0.414
trunk -0.582
weight -0.807
length -0.796
displacement -0.706
. mat C = r(C)
. mat li C
C[5,1]
mpg
headroom -.41380299
trunk -.58158503
weight -.80717486
length -.79577944
displacement -.70564259
. pca headroom trunk weight length displacement
Principal components/correlation Number of obs = 74
Number of comp. = 5
Trace = 5
Rotation: (unrotated = principal) Rho = 1.0000
--------------------------------------------------------------------------
Component | Eigenvalue Difference Proportion Cumulative
-------------+------------------------------------------------------------
Comp1 | 3.76201 3.026 0.7524 0.7524
Comp2 | .736006 .427915 0.1472 0.8996
Comp3 | .308091 .155465 0.0616 0.9612
Comp4 | .152627 .111357 0.0305 0.9917
Comp5 | .0412693 . 0.0083 1.0000
--------------------------------------------------------------------------
Principal components (eigenvectors)
----------------------------------------------------------------
Variable | Comp1 Comp2 Comp3 Comp4 Comp5
-------------+--------------------------------------------------
headroom | 0.3587 0.7640 0.5224 -0.1209 0.0130
trunk | 0.4334 0.3665 -0.7676 0.2914 0.0612
weight | 0.4842 -0.3329 0.0737 -0.2669 0.7603
length | 0.4863 -0.2372 -0.1050 -0.5745 -0.6051
displacement | 0.4610 -0.3390 0.3484 0.7065 -0.2279
----------------------------------------------------------------
---------------------------
Variable | Unexplained
-------------+-------------
headroom | 0
trunk | 0
weight | 0
length | 0
displacement | 0
---------------------------
. predict PC1-PC5
(score assumed)
Scoring coefficients
sum of squares(column-loading) = 1
----------------------------------------------------------------
Variable | Comp1 Comp2 Comp3 Comp4 Comp5
-------------+--------------------------------------------------
headroom | 0.3587 0.7640 0.5224 -0.1209 0.0130
trunk | 0.4334 0.3665 -0.7676 0.2914 0.0612
weight | 0.4842 -0.3329 0.0737 -0.2669 0.7603
length | 0.4863 -0.2372 -0.1050 -0.5745 -0.6051
displacement | 0.4610 -0.3390 0.3484 0.7065 -0.2279
----------------------------------------------------------------
. cpcorr headroom trunk weight length displacement \ PC?
(obs=74)
PC1 PC2 PC3 PC4 PC5
headroom 0.6958 0.6554 0.2900 -0.0472 0.0026
trunk 0.8405 0.3144 -0.4261 0.1138 0.0124
weight 0.9392 -0.2856 0.0409 -0.1043 0.1545
length 0.9432 -0.2035 -0.0583 -0.2245 -0.1229
displacement 0.8942 -0.2909 0.1934 0.2760 -0.0463
. cpcorr headroom trunk weight length displacement \ PC?, format(%3.2f)
(obs=74)
PC1 PC2 PC3 PC4 PC5
headroom 0.70 0.66 0.29 -0.05 0.00
trunk 0.84 0.31 -0.43 0.11 0.01
weight 0.94 -0.29 0.04 -0.10 0.15
length 0.94 -0.20 -0.06 -0.22 -0.12
displacement 0.89 -0.29 0.19 0.28 -0.05
Code:
sysuse auto, clear
mat ktau = J(5, 1, .)
local Y mpg
local nY : word count `Y'
local X headroom trunk weight length displacement
local nX : word count `X'
forval i = 1/`nX' {
local x : word `i' of `X'
forval j = 1/`nY' {
local y : word `j' of `Y'
quietly ktau `x' `y'
matrix ktau[`i', `j'] = r(tau_a)
}
}
matrix colnames ktau = `Y'
matrix rownames ktau = `X'
mat li ktau, format(%4.3f)
ktau[5,1]
mpg
headroom -0.305
trunk -0.451
weight -0.686
length -0.633
displacement -0.594
