|
DM SAS macros included |
Required SAS MODULES |
Sample OUTPUT (PDF) |
| 1)
EXCELSAS : - Converting PC file types to SAS data
|
SAS/BASE & SAS/ACCESS
(PC files)
SAS/LE
Compatibility:
PROC IMPORT is not
included in SAS/LE. This MACRO won't work to convert
excel file type. But,
SAS Enterprise guide (EG)
can import many different PC file types directly.
OR convert excel file
to TAB delimited file and run the EXCELSAS macro and select the file
type to TAB. |
EXCELSAS OUTPUT |
| 2)
RANSPLIT:- Creating Training,
validation, and test data sets |
SAS/BASE & SAS/STAT (Box-plot)
SAS/LE
Compatibility:
Fully functional in SAS/LE |
RANSPLIT OUTPUT |
| 3)
FREQ:- Exploring one- two- and three-way categorical
data |
SAS/BASE & SAS/GRAPH
SAS/LE
Compatibility:
Fully functional in SAS/LE |
FREQ OUTPUT 1
FREQ OUTPUT 11 |
| 4)UNIVAR: Exploring continuous variable (s) -with
and without a group variable |
SAS/BASE , SAS/STAT, SAS/GRAPH, & SAS/QC
(Comparative histogram)
SAS/LE
Compatibility:
Fully functional in SAS/LE |
UNIVAR OUTPUT 1
UNIVAR OUTPUT
11 |
| 5)FACTOR: Principal component and exploratory factor
analysis - exploration, checking for multivariate normality, Bi-plot
display |
SAS/BASE , SAS/STAT, SAS/GRAPH, & SAS/IML
(Multivariate normality check)
SAS/LE
Compatibility:
Fully functional in SAS/LE, (
Except for multivarite
normality stats using SAS/IML is not avilable) |
FACTOR OUTPUT 1
FACTOR OUTPUT II |
| 6)DISJCLUS:
Dis-Joint cluster analysis, data exploration - checking for
multivariate normality, canonical discriminant analysis of cluster
groupings and Bi-plot display canonical discriminant functions |
SAS/BASE , SAS/STAT, SAS/GRAPH, & SAS/IML
(Multivariate normality check)
SAS/LE
Compatibility:
Fully functional in SAS/LE, (
Except for multivarite
normality stats using SAS/IML
) |
DISJCLUS OUTPUT 1
DISJCLUS
OUTPUT II |
| 7)
REGDIAG: Multiple
linear regression modeling (REG & GLM) , diagnostic partial plots,
interaction detection plots, violation assumption plots, model selection
plots and model validation plots and analyses. |
SAS/BASE , SAS/STAT, and SAS/GRAPH
SAS/LE
Compatibility:
Fully functional in SAS/LE |
REGDIAG OUTPUT 1 |
| 8)RSCORE:
Predicting and scoring an independent dataset |
SAS/BASE , SAS/STAT, and SAS/GRAPH
SAS/LE
Compatibility:
Limited functionalityin SAS/LE, (
Scoring a new data is only
possible if the data contain the observed response value. |
RSCORE OUTPUT |
| 9)
LIFT: IF-THEN or
sensitivity analysis in multiple linear and logistic regression model |
SAS/BASE , SAS/STAT, and SAS/GRAPH
SAS/LE
Compatibility:
Fully functional in SAS/LE |
LIFT OUTPUT |
| 10)LOGISTIC:Logistic
regression modeling (PROC LOGISTIC) , diagnostic partial logit plots,
interaction detection plots, violation assumption plots, model
selection, ROC curve, and model validation plots and analyses. |
SAS/BASE , SAS/STAT, and SAS/GRAPH
SAS/LE
Compatibility:
Fully functional in SAS/LE |
LOGISTIC1 OUTPUT
LOGISTIC2
OUTPUT |
| 11)
LSCORE: Predicting and
scoring of an independent data based on estimated logistic regression
model |
SAS/BASE , SAS/STAT, and SAS/GRAPH
SAS/LE
Compatibility:
Limited functionalityin SAS/LE, (
Scoring a new data is only
possible if the data contain the observed response value. |
LSCORE OUTPUT |
| 12)
DISCRIM: Stepwise,
Canonical, Parametric and non-parametric Discrimaint analysis, data
exploration, checking for multivariate normality, Bi-plot display |
SAS/BASE , SAS/STAT, SAS/GRAPH, & SAS/IML
(Multivariate normality check)
SAS/LE
Compatibility:
Fully functional in SAS/LE, (
Except for multivarite
normality stats using SAS/IML ) |
DISCRIM OUTPUT1
DISCRIM
OUTPUT1I |
| 13)
CHAID: Data exploration
and Decision tree analysis |
SAS/BASE , SAS/IML, SAS/GRAPH, & SAS/OR
(NETWORK diagram) SAS/LE
Compatibility: SAS/IML
is not included in SAS/LE. CHAIDMACRO won't work in LE |
CHAID OUTPUT |