# st: discriminant analysis output interpretation

 From Yan elaine Li To statalist@hsphsun2.harvard.edu Subject st: discriminant analysis output interpretation Date Wed, 12 May 2004 21:37:45 -0700 (PDT)

```Dear Statalisters,

can anyone help with the following user's question?  Thanks!

Yan Li

---------- Forwarded message ----------

the name of the function is daoneway and it performs discriminant analysis

it performs discriminant analysis and the output gives me four sets of
functions:

Unstandardized canonical discriminant function coefficients

standardized canonical discriminant function coefficients

Canonical discriminant structure matrix

Group means on canonical discriminant functions

My question is with interpretation of those functions.
i am unsure which of these functions i should use to assign a new data
point. in other words, how do i use this output to do what discriminant
anaysis is supposed to do,that is, assign a new point to a group.

an example output is below:

. daoneway x1 x2 x3 x4, by (y)

One-way Disciminant Function Analysis

Observations = 20
Variables    = 4
Groups       = 3

Pct of   Cum  Canonical  After  Wilks'
Fcn Eigenvalue Variance  Pct     Corr      Fcn  Lambda  Chi-square
df	P-val
> ue
0  0.31918    17.701	8	0.02
> 36
1    1.2262   75.07  75.07    0.7422     1  0.71057     5.296	3
0.15
> 14
2    0.4073   24.93 100.00    0.5380

Unstandardized canonical discriminant function coefficients

func1    func2
x1  48.1854  -2.6310
x2  -0.3697   4.0740
x3   0.1082   0.0960
x4  -0.0100   0.0890
_cons  -0.6798  -3.7256

Standardized canonical discriminant function coefficients

func1    func2
x1   0.8634  -0.0471
x2  -0.0812   0.8951
x3   0.5570   0.4941
x4  -0.0224   0.1992

Canonical discriminant structure matrix

func1    func2
x1   0.8238  -0.2015
x2  -0.2746   0.8472
x3   0.4871   0.4390
x4   0.2206   0.0768

Group means on canonical discriminant functions

func1    func2
y-1  -0.7413  -0.9253
y-2  -0.3983   0.4802
y-3   2.0220  -0.1639

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```