Copyright © 2006, SAS Institute Inc. All rights reserved. A Sampler of What's New in Base SAS...

A Sampler of What's New in Base SAS 9.2Jason.Secosky@sas.com

Caramel, Nut, or Truffle?

DATA Step• User-Written Functions

• Data Set Lists

• Current Input Data Set Name

• INPUT with String Delimiter

PROC SQL• Un-PUT

• Faster DISTINCT

PROC SORT Linguistic Ordering

Encrypted Macro Storage

Caramel, Nut, or Truffle?

DATA Step• User-Written Functions

• Data Set Lists

• Current Input Data Set Name

• INPUT with String Delimiter

PROC SQL• Un-PUT

• Faster DISTINCT

PROC SORT Linguistic Ordering

Wrap in a Macro

%macro study_day(start, event, study_day);

n = &event - &start;

if n >= 0 then

&study_day = n + 1;

&study_day = n;

%mend;

Bug in the Macro

data results;

set trial_data;

n = sum(of visits_jan--visits_dec);

%study_day(start, event, study_day)

if n > 5 then

Bug in the Macro

data results;

set trial_data;

n = event – start;

if n >= 0 then

study_day = n + 1;

study_day = n;

if n > 5 then

Protected by a Function

data results;

set trial_data;

study_day = study_day(start, event);

if n > 5 then

Study Day Function

proc fcmp outlib=sasuser.funcs.trial;

function study_day(start, event);

n = event – start;

if n >= 0 then

return(n + 1);

return(n);

endsub;

Data Set Lists

#1 on 2006 SASware Ballot

Read Like-Named Data Sets

Syntax borrowed from DROP and KEEP

Read a Bunch of Data Sets

data out;

set a1 a2 a3 a4 a5 a6 a7 a8 a9 a10

a11 a12 a13 a14 a15 a16 a17 a18

a19 a20 a21 a22 a23 a24 a25 a26

a27 a28 a29 a30 a31 a32 a33 a34

a35 a36 a37 a38 a39 a40 a41 a42

a43 a44 a45 a46 a47 a48 a49 a50;

Data Set Lists

data out;

set a1-a50(pw=pass123);

data out;

set sales.jan:;

String Delimiter

"Top 10" SASware Ballot Item

Difficult to parse multi-character delimiters

INFILE DLMSTR=

Works like DLM=

String Delimiter

data bad_guys;

length date time ip $ 16;

infile datalines dlm=' ' dsd;

input date time ip;

datalines;

"20APR2007" "12:56:46" "146.16.1.23"

String Delimiter

data bad_guys;

length date time ip $ 16;

infile datalines dlmstr=' ' dsd;

input date time ip;

datalines;

"20APR2007" "12:56:46" "146.16.1.23"

SET with INDSNAME=

"Top 10" SASware Ballot Item

Place current data set name into a variable

SET with INDSNAME = var

SET with INDSNAME=

data results;

set gas_price_option

gas_rbid_option

coal_price_forward

coal_rbid_forward

indsname = cur_dataset;

DATA Step JavaObj

"Compute" methods written in Java

Call from DATA Step with JavaObj

Methods like Java Native Interface (JNI)

Uses dot syntax

Production in SAS 9.2

Experimental in SAS 9.1.3

data results;

declare javaobj jo("myclass");

jo.callDoubleMethod("compute",

x, y, z, res);

DATA Step JavaObj

DATA Step IN Operator

IN operator used simple array search

Now, IN searches a binary tree

9 12 14 17 19 23 50 54 67 72 76

data _null_;

set customers;

if state in ('NC', 'SC', 'GA',

'TN', 'FL');

DATA Step IN Operator

Proc SQL Today

proc format;

value udfmt 1-3='small'

4-6='medium'

7-9='large';

proc sql;

select style as SmallStyles

from oracle.clothes

where put(size, udfmt.) = 'small';

Proc SQL Today

SAS® Session

select style as SmallStylesfrom oracle.clotheswhere put(size, udfmt.) = 'small';

SAS/Access Engine

select style,size from clothes

Proc SQL Today

SAS® Session

SAS/Access Engine

select style,size from clothes

SAS® Session

Proc SQL Tomorrow

SAS/Access Engine

select style as SmallStylesfrom oracle.clotheswhere (1 <= size and size <= 3);

SAS® Session

Proc SQL Tomorrow

SAS/Access Engine

select style as SmallStylesfrom oracle.clotheswhere (1 <= size and size <= 3);

Proc SQL Today

proc sql;

select name

from oracle.employees

where put(bday, date5.) = '31OCT';

Proc SQL Tomorrow

proc sql;

select name

from oracle.employees

where month(bday) = 10 and

day(bday) = 31;

Proc SQL Today

proc format;

value nudfmt 0 = 'RED'

1 = 'NOTRED'

2 = 'GREEN'

other = 'BLACK';

proc sql;

select * from oracle.data where

put(color, nudfmt.) contains "RED";

Proc SQL Tomorrow

proc format;

value nudfmt 0 = 'RED'

1 = 'NOTRED'

2 = 'GREEN'

other = 'BLACK';

proc sql;

select * from oracle.data where

color in (0,1);

PROC SQL

Faster: SELECT DISTINCT

Faster: COUNT(DISTINCT var)

Hash Join Performance

Row Counts from DBMSes

SELECT DISTINCT

proc sql;

create table distinct_values as

select distinct month, year

from customer_info;

COUNT DISTINCT

proc sql;

create table count_distinct as

select count(distinct customerid)

from customer_info;

Linguistic Sorting

Obs Binary Lowfirst Linguistic

1 Aaron aardvark aardvark

2 Aztec azimuth Aaron

3 Zeus Aaron azimuth

4 aardvark Aztec Aztec

5 azimuth zebra zebra

6 zebra Zeus Zeus

Linguistic Sorting

proc sort data=names

out=sorted

sortseq=linguistic;

by name;

Numeric_Collation=ON

Obs Binary Linguistic

1 0123 Main St. Apt #12 123 Main St. Apt #1

%macro mymac(x, y) / store secure;

%mend;

More Details

Search support.sas.com for:

Papers:

"User-Written DATA Step Functions"

"Creating Order out of Character Chaos"

Copyright © 2006, SAS Institute Inc. All rights reserved. A Sampler of What's New in Base SAS...

Documents

CONTRAST、およびESTIMATE ... - sas.com · また、sas 出力は全てオリジナルと同様に英語版sas の出力です。日本語版sas でプロ日本語版SAS でプロ

Copyright © 2011, SAS Institute Inc. All rights reserved. SAS ® Curriculum Pathways ® Ralph Moore ralph.moore@sas.com

Open Source Integration - SAS...Title: Open Source Integration Author: Tuba Islam;Dan.Zaratsian@sas.com Created Date: 9/24/2018 9:43:06 AM

Machine learning and sas · Machine Learning with SAS • Introduction to ML –concepts, motivation, possibilities ... KAARE.BRANDT@SAS.COM +45 51387884. Title: Machine learning

Datenanalyse mit SAS - sas.com · Diese Einführung in die Datenanalyse mit SAS enthält einen Überblick über die analytische Funktiona-lität der SAS Plattform. Sie richtet sich

Practical Bayesian Computation using SAS®Practical Bayesian Computation using SASR Fang Chen SAS Institute Inc. fangk.chen@sas.com ASA Conference on Statistical Practices February

SAS Information Delivery Portal from sas.com

SAS VIYA - sas.com · ARQUITECTURA MULTI-VENDOR REQUERIMIENTOS DEL MERCADO SAS 9 MVS Windows HP-UX AIX Solaris SunOS Linux Tru64 IRIX OS/2. Company Confidential - For Internal Use

GOVERNMENT ANALYTICS LEADERSHIP FORUM - sas.com · GOVERNMENT ANALYTICS LEADERSHIP FORUM SAS Canada & The Institute of Public Administration of Canada April 26 | The Shaw Centre |

MATERIAL SAMPLER BOOK - SAS Industries · MATERIAL SAMPLER BOOK 07 Sponge Rubbers are great gasketing materials for applications requiring soft, compressible material. Bisco™ offers

MATERIAL SAMPLER BOOK - SAS Industries Inc

Copyright © 2006, SAS Institute Inc. All rights reserved. Proc FCMP – Data Step Functions Denise.Poll@sas.com

Update From SAS Group Presentati… · Build Online Community with SAS blogs, LinkedIn Groups, etc. Matt Joyce Customer Loyalty Executive Tel: (403) 802 4474 Cell: (587)215 2529 Matthew.Joyce@sas.com

Analyzing Enhanced Oil Recovery - sas.com Group Presentation… · Gambit Consulting Ltd Presentation for Calgary SAS Users Group 08 Nov. 2013 Analyzing Enhanced Oil Recovery using

SAS Visual Analytics 8 · Whats’snew in SAS Visual Analytics 8.4 niklas.schulman@sas.com. High Level Comparison of VA Releases HTML5 Viewer Flash Viewer Designer Explorer 7.4 7.5

SAS+ SAP = SANT - sas.com · PDF file• Partnership Announced at SAP TechEd ... SAS for SAP HANA Roadmap 2013 2014 2014 2014 H2 Q2 Q3 Q4 ... SAS+ SAP = SANT

EL VALOR DE LA PALABRA DEL CLIENTE - SAS TEXT MINING · EL VALOR DE LA PALABRA DEL CLIENTE - SAS TEXT MINING rosanamac.lean@sas.com. ... documents, audio, video and correspondence

FORAGE - tec-sas.com

4 27.-28.5.201 O OULUN KURSSI mainos SAS koodaus 2pv... · Little SAS Book -opus omaksi. ! LISÄTIETOJA: Kurssin opetuskieli on suomi. Lisätiedot: SAS juha.jokinen@sas.com! KURSSI

Performance Tuning your SAS Environment · 2016-03-11 · Performance Tuning your SAS Environment (Hints and Tips) Andrew Gadsby SAS UK Customer Loyalty Team andrew.gadsby@sas.com