Friday, May 30, 2014

SAS Resources and blogs

1. 人大
http://bbs.pinggu.org/forum-219-1.html

Proc Lifetest and PROC PHREG

1. Tips for Creating Oncologic Efficacy Summary Tables using PROC  LIFETEST and PROC PHREG
http://www.lexjansen.com/pharmasug/2010/ad/ad03.pdf

Tuesday, May 27, 2014

About ODS - Output Delivery System

1. Output Delivery  System Tip Sheet
ODS allows you to format your reports in various formats such as HTML, PDF, RTF, Microsoft Excel, and many others. It enables you to customize your reports by selecting only the results you want to see. It even lets you apply styles to your reports including many supplied by SAS® or those that you create.

http://support.sas.com/rnd/base/ods/scratch/ods-tips.pdf

2. Get SAS proc output object by ODS trace on
http://www.ats.ucla.edu/stat/sas/faq/odsexample.htm

3. Output PDF file
http://www2.sas.com/proceedings/sugi30/172-30.pdf

4. Advanced Mobile Reporting with the ODS EPUB3 Destination - iPhone, iPad
http://support.sas.com/resources/papers/proceedings14/SAS339-2014.pdf

5. Customize SAS® Output Files Using SAS Output Delivery System
http://www.lexjansen.com/nesug/nesug08/cc/cc02.pdf

6. SAS ODS output
https://support.sas.com/rnd/base/ods/odsmarkup/

  • CSV, CSVALL, CSVBYLINE
  • ExcelXP
  • HTML with graph bars
  • htmlpanel
  • htmlscroll
  • jQueryMobile
  • LaTeX
  • MSOffice2K_x
  • SQL
  • Super Map!
  • tableeditor

Saturday, May 24, 2014

Get into the door of DA Career

一、掌握基础、更新知识。

基本技术怎么强调都不过分。这里的术更多是(计算机、统计知识), 多年做数据分析、数据挖掘的经历来看、以及业界朋友的交流来看,这点大家深有感触的。

数据库查询—SQL

数据分析师在计算机的层面的技能要求较低,主要是会SQL,因为这里解决一个数据提取的问题。有机会可以去逛逛一些专业的数据论坛,学习一些SQL技巧、新的函数,对你工作效率的提高是很有帮助的。

统计知识与数据挖掘

你要掌握基础的、成熟的数据建模方法、数据挖掘方法。例如:多元统计:回归分析、因子分析、离散等,数据挖掘中的:决策树、聚类、关联规则、神经网络等。但是还是应该关注一些博客、论坛中大家对于最新方法的介绍,或者是对老方法的新运用,不断更新自己知识,才能跟上时代,也许你工作中根本不会用到,但是未来呢?

行业知识

如果数据不结合具体的行业、业务知识,数据就是一堆数字,不代表任何东西。是冷冰冰,是不会产生任何价值的,数据驱动营销、提高科学决策一切都是空的。

一名数据分析师,一定要对所在行业知识、业务知识有深入的了解。例如:看到某个数据,你首先必须要知道,这个数据的统计口径是什么?是如何取出来的?这个数据在这个行业, 在相应的业务是在哪个环节是产生的?数值的代表业务发生了什么(背景是什么)?对于A部门来说,本月新会员有10万,10万好还是不好呢?先问问上面的这个问题:

四、业务、行业、商业知识。

当你掌握好前面的基本知识和一些技巧性东西的时候,你应该在业务、行业、商业知识的学习与积累上了。

这个放在最后,不是不重要,而且非常重要,如果前面三点是决定你能否进入这个行业,那么这则是你进入这个行业后,能否成功的最根本的因素。 数据与具体行业知识的关系,比作池塘中鱼与水的关系一点都不过分,数据(鱼)离开了行业、业务背景(水)是死的,是不可能是“活”。而没有“鱼”的水,更像是“死”水,你去根本不知道看什么(方向在哪)。

如何提高业务知识,特别是没有相关背景的同学。很简单,我总结了几点:

1、多向业务部门的同事请教,多沟通。多向他们请教,数据分析师与业务部门没有利益冲突,而更向是共生体,所以如果你态度好,相信业务部门的同事也很愿意把他们知道的告诉你。
2、永远不要忘记了google大神,定制一些行业的关键字,每天都先看看定制的邮件。
3、每天有空去浏览行业相关的网站。看看行业都发生了什么,主要竞争对手或者相关行业都发展什么大事,把这些大事与你公司的业务,数据结合起来。
4、有机会走向一线,多向一线的客户沟通,这才是最根本的。

以上几点更多我自己的一些心得的总结。希望对新进的朋友有帮助,数据分析行业绝对是一个朝阳行业,特别是互联网的不断发展,一个不谈数据的公司根本不叫互联网公司,数据分析师已经成为一个互联网公司必备的职位了。

SAS China
----------------------
数据时代来了,无穷的数据吞噬了整个世界,无厘头的数据阻碍和困扰世界文明前进发展,正在人们不知所措时SAS这把利器征服了大数据,古人云:工欲善其事必先利其器。想要拯救和主宰这个世界必然要学会利用这把利器。数学中国《第二期SAS实训》名师坐镇亲自教会大家如何使用SAS 这把利器来主宰这个世界5 @2 C# G, {* L* H8 o" W& T" g
         
      主讲教师:郭海兵   B1 A7 z, |0 U& c; j
       教师简介:中国人民大学统计学博士,淮海工学院教师
擅长数据分析和统计建模,熟悉常用的统计软件;长期担任SAS公司兼职讲师,主讲OR(最优化)模块、分类数据分析、统计12等课程;拥有丰富的数学建模指导经验,带队学生参加美国数学建模竞赛,多次获得一等奖和二等奖。
       PS :之前很多同学问我SAS在数学建模中如何应用。咱们的主讲老师也是数学建模指导教师在培训时大家可以和老师多多互动交流学习经验。" k( V2 N2 W% |# E% w1 \& d4 x
      招生对象:SAS爱好者、数学建模爱好者、大数据爱好者、SAS教师
      课程时间:2014年5月25日——2014年6月8日  . d& v2 R/ ?, w. h$ q& V) L0 }7 Y* G
    课程内容设定:
         
      第一章、SAS系统简介
    第二章、SAS/INSIGHT模块
    第三章、DATA步与数据处
    第四章、常用PROC语句
    第五章、数据描述与MEANS过程
    第六章、单变量分析与UNIVARIATE过程
    第七章、统计图表与SAS过程
    第八章、列联表数据与FREQ过程
    第九章、回归分析与相关分析
    第十章、方差分析与ANOVA过程
    第十一章、统计诊断模型选择(备选)
   第十二章、分类数据与LOGISTIC过程(备选) 
         
         本次课程内容的设定主要针对了SAS零基础的同学和想要参加  中国高校SAS数据分析大赛  做好提前热身准备。参加数学建模的同学在比赛中大数据的题也可以使用SAS来完成。必将创造一大亮点。
      A% N% f; s( O" [. n
       上课方式:YY语音+桌面共享+课件讲义+课后视频 (错过上课的同学可以无限次观看视频)
      报名费用:
         
报名人数单人报名三人团报三人以上团报
费用(全部培训内容)158元每人150元每人145元每人
注:老学员给予150元的优惠价格(有学号的)

         报名方式: 点击页面中的我要参加 ,完善信息。按缴费步骤完成缴费。  

                              或联系工作人员乔叶  QQ 1470495151                         $ f3 v( A; t. d( K8 H3 k
      缴费方式:
            1> 数学中国淘宝店缴费:
数学中国淘宝店(点击淘宝缴费) 完成交易后将交易号 或交易成功截图及姓名 发到 ceo@madio.cn
            2>支付宝转账,帐号为:ilikenba@263.net  ,转账成功后,将支付宝交易成功截图和交易号 和姓名发到 ceo@madio.cn   U, e  t4 s/ q& ^
            3> 银行汇款:请通过银行柜面或者ATM机或网银将报名费汇至中国工商银行内蒙古分行营业部明珠支行,账号6222 3115 3512 7013,户名:马壮,汇款后请将”汇款后请将”汇款时间+汇款金额“发邮件到ceo@madio.cn说明”。有条件者拍照片或截图发到邮箱。

SAS Users Group International (SUGI) , SAS Conference Proceedings

1. SUGI 30 Proceedings
http://www2.sas.com/proceedings/sugi30/toc.html

2. SAS Conference Proceedings
SAS Conference Proceedings: SAS Global Forum 2012
SASGF 2014   SASGF 2013   SASGF 2012   SASGF 2011   SASGF 2010   SASGF 2009   SASGF 2008   SASGF 2007   SUGI 31   SUGI 30   SUGI 29   SUGI 28   SUGI 27   SUGI 26   SUGI 25   SUGI 24   SUGI 23   SUGI 22   SUGI 21   SUGI 20   SUGI 19   SUGI 18   SUGI 17   SUGI 16   SUGI 15  SUGI 14   SUGI 13   SUGI 12   SUGI 11   SUGI 10   SUGI 1984   SUGI 1983   SUGI 1982   SUGI 1981   SUGI 1980   SUGI 1979   SUGI 1978   SUGI 1977   SAS.ONE 1976  

3.  SAS global Forum
http://support.sas.com/events/sasglobalforum/previous/online.html

A complete ref --- Introduction to SAS, Data Analysis with Examples

1. Introduction to SAS
https://www.stat.wisc.edu/~yandell/software/sas/#book

2. Data Analysis Exampleshttp://www.ats.ucla.edu/stat/dae/

Friday, May 23, 2014

Non-parametric Tests - When, Why, and How

1. An Overview of Non-parametric Tests in SAS -- When, Why, and How
http://analytics.ncsu.edu/sesug/2004/TU04-Pappas.pdf


Wednesday, May 21, 2014

LOGISTIC Regression analysis, Proc GLM, Proc Reg

1. Concept
http://baike.baidu.com/view/145440.htm
回归分析(regression analysis)是确定两种或两种以上变数间相互依赖的定量关系的一种统计分析方法。运用十分广泛,回归分析按照涉及的自变量的多少,可分为一元回归分析和多元回归分析;按照自变量和因变量之间的关系类型,可分为线性回归分析和非线性回归分析。如果在回归分析中,只包括一个自变量和一个因变量,且二者的关系可用一条直线近似表示,这种回归分析称为一元线性回归分析。如果回归分析中包括两个或两个以上的自变量,且因变量和自变量之间是线性关系,则称为多元线性回归分析。
回归分析(英语:Regression Analysis)是一种统计学上分析数据的方法,目的在于了解两个或多个变量间是否相关、相关方向与强度,并建立数学模型以便观察特定变量来预测研究者感兴趣的变量。

      Logistic regression (逻辑回归) 概述
http://hi.baidu.com/hehehehello/item/40025c33d7d9b7b9633aff87



2. Proc GLM Explained
http://screamyao.wordpress.com/2010/09/30/sas-proc-glm-explained/

3
https://support.sas.com/documentation/cdl/en/statugglm/61789/PDF/default/statugglm.pdf

4.Lesson 13:  Proc Reg
http://galsterhome.com/stats/Tutorial/SAS13.htm

5.The REG Procedure
http://www.math.wpi.edu/saspdf/stat/chap55.pdf

6. How do I interpret odds ratios in logistic regression
http://www.ats.ucla.edu/stat/sas/faq/oratio.htm

7. Multinomial and ordinal logistic regression using PROC LOGISTIC
http://www.nesug.org/proceedings/nesug05/an/an2.pdf

8. A Tutorial on PROC LOGISTIC
http://www.mwsug.org/proceedings/2013/RX/MWSUG-2013-RX08.pdf
a.. The meaning of Relative Risk & Odds Ration
b. Clinical Study Design: cross-sectional, cohort (prospective), and case-control (retrospective) study.


9. Illustrative Logistic Regression Examples using PROC LOGISTIC: New Features InsuranceSAS9.2
http://www.lexjansen.com/pharmasug/2009/sp/sp03.pdf

10. Linear Models in SAS
https://www.stat.wisc.edu/~yandell/software/sas/linmod.html#anova


SAS/STAT Procedures A-Z

1. SAS/STAT Procedures A-Z

http://support.sas.com/rnd/app/stat/procedures/Procedures.html

http://www.iasri.res.in/sscnars/sas_manual/1-SAS%20an%20overview%20for%20statistical%20procedures.pdf

PROC CONTENTS: Handle multiple dataset automatically

1. Using the Contents of PROC CONTENTS to Perform Multiple Operations Across a SAS®
Data Library
http://www2.sas.com/proceedings/sugi27/p084-27.pdf

Tuesday, May 20, 2014

Hypo Test

1. Comparing Two Groups with PROC TTEST
http://www.sas.com/offices/europe/belux/pdf/academic/ttest.pdf

2. Example SAS code for a two-sample T-Test:

http://facweb.cs.depaul.edu/cmiller/it223/ttest.html


Proc UNIVARIATE

1. The UNIVARIATE Procedure
http://www.math.wpi.edu/saspdf/proc/c41.pdf

2. Guido’s Guide to PROC UNIVARIATE: A Tutorial for SAS® Users
http://www.nesug.org/Proceedings/nesug09/sa/sa07.pdf

Proc Corr: Correlation

1. The CORR Procedure
http://www.math.wpi.edu/saspdf/proc/c12.pdf

2. Getting Correlations Using PROC CORR
http://www.stat.wvu.edu/~abilling/STAT521_ProcCorrPlot.pdf

Friday, May 16, 2014

Proc Compare, Transpose, SYMPUT

1. PROC COMPARE – Worth Another Look!
http://support.sas.com/resources/papers/proceedings10/149-2010.pdf

2. A Row is a Row is a Row, or is it? Get Comfortable with Transposing your Data
http://support.sas.com/resources/papers/proceedings09/033-2009.pdf

3. SYMPLIFY your Data Set Transposition with SYMPUT, and Make it Data-Driven Too!
http://analytics.ncsu.edu/sesug/2006/CC05_06.PDF

Tuesday, May 13, 2014

Data Analyst Skills Set

1.  SQL, Database
Oracle PL/SQL Development Experience with Strong ETL experience.
Good understanding of Relational Database concepts.
Experienced with writing stored procedures, functions and triggers.
Experience working with TOAD/SQL Developer.

2. Excel, VBA
Exceptionally strong Excel skills (Microsoft certified preferred)
Able to decipher, interpret, troubleshoot, and edit complex Excel formulas including (but certainly not limited to):
VLOOKUPs
OFFSET references
Nested IF Statements

Advanced Excel macro development including:
Writing VBA code specifically for Excel and the functions associated with the program
Troubleshooting and resolving bugs in existing VBA code
Experience with editing MS Office Ribbons using XML

Monday, May 12, 2014

EXCEL + HTML + CSV file from and to SAS dataset

1. Excellent Ways of Exporting SAS Data to Excel
http://www.nesug.org/proceedings/nesug04/io/io09.pdf

Sunday, May 11, 2014

Google Analytics API and SAS

1. Bridging the Gap between the Google Analytics API and SAS
http://support.sas.com/resources/papers/proceedings10/049-2010.pdf

2. Building Business Intelligence with SAS and Google Analytics
http://support.sas.com/resources/papers/proceedings12/010-2012.pdf

3. google analytics certification
https://support.google.com/analytics/answer/3424287?hl=zh-Hans
http://blog.sina.com.cn/s/blog_79ccab6801019sox.html
http://blog.sina.com.cn/s/blog_79ccab6801019srv.html
http://www.an7.me/archives/897
https://support.google.com/analytics/answer/4553001?hl=zh-Hans

Thursday, May 8, 2014

Clinical Trial - Statistics

1. Medical Statistics: Clinical Trials
http://www.nickfieller.staff.shef.ac.uk/sheff-only/clinical.pdf

2. SAS programming in the pharmaceutical industry
http://www.planta.cn/forum/files_planta/sas_programming_in_the_pharmaceutical_industry_119.pdf

3. Statistical Principles of Clinical Trials
http://www4.stat.ncsu.edu/~dzhang2/st520/520notes.pdf

Association Test: Proc Freq , Chi-Square Tests

1. Guido’s Guide to PROC FREQ – A Tutorial for Beginners
http://www.nesug.org/proceedings/nesug07/ff/ff07.pdf

2. Glenn Walker book “Common Statistical Methods for Clinical Research with SAS® Examples
http://www.sas.com/storefront/aux/en/spcommonstat/62004_excerpt.pdf

3. Answering the Right Question with the Right PROC

4. Statistics I: Introduction to ANOVA, Regression, and Logistic Regression
http://gendocs.ru/docs/18/17094/conv_1/file1.pdf

5. 统计检验之——卡方检验; 北京宏志中学 徐德前
http://education.ti.com/sites/CHINA/downloads/pdf/chi_square_tests_xudeqian.pdf

6. PROC FREQ IS MORE THAN JUST SIMPLY GENERATING A 2-BY-2 TABLE
http://www.lexjansen.com/wuss/2012/52.pdf

7. THE POWER OF PROC FORMAT
http://www.ats.ucla.edu/stat/sas/library/nesug00/bt3001.pdf

8.PROC FORMAT in Action
http://www2.sas.com/proceedings/sugi27/p056-27.pdf


9. PROC FORMAT – Not Just Another Pretty Face
http://www2.sas.com/proceedings/sugi30/001-30.pdf

Wednesday, May 7, 2014

Building Business Intelligence with SAS® and Google Analytics

http://support.sas.com/resources/papers/proceedings12/010-2012.pdf

http://support.sas.com/resources/papers/proceedings10/049-2010.pdf

Forecast

1. Proc Forecast

http://www.okstate.edu/sas/v8/saspdf/ets/chap12.pdf
Syntax: pg 18
STEPAR Method: not optimal but close to, and computational inexpensive

TREND=2=linear model
NLAGS>=3, 13
SLENTRY= option=significant at the level=0.20
SLSTAY=0.05=significance level 
 OUT= data set
Missing values are tolerated in the series
METHOD=EXPO


SAS functions

1. WORKING WITH SAS DATE AND TIME FUNCTIONS
http://www.ats.ucla.edu/stat/sas/library/nesug00/bt3007.pdf

2. SAS functions by example
http://www.sas.com/storefront/aux/en/spfunctionxexample/62857_excerpt.pdf

Proc SQL and equievlent in Data Step, Fuzzy match

1. TOP 10 FUNCTIONS FOR THE SQL PROCEDURE IN SAS

http://www.sasanalysis.com/2011/01/top-10-most-powerful-functions-for-proc.html


2. Fuzzy match
http://support.sas.com/documentation/cdl/en/lrdict/64316/HTML/default/viewer.htm#a000245949.htm
http://www2.sas.com/proceedings/sugi25/25/cc/25p086.pdf
http://www.nesug.org/Proceedings/nesug11/ap/ap07.pdf

3. PROC SQL for DATA Step Die-hards
http://www2.sas.com/proceedings/forum2008/185-2008.pdf

4. PROC SQL: From SELECT to Pass-Through SQL
http://www.scsug.org/SCSUGProceedings/2010/Schacherer_2/PROC_SQL_%20From_SELECT_to_Pass-Through_SQL.pdf

5. 集合操作:inner join、left join、right join、full outer join、union、union all
http://blog.diyiye.com/?post=10

6. Queries, Joins, and WHERE Clauses, Oh My!! Demystifying PROC SQL
http://analytics.ncsu.edu/sesug/2012/HW-06.pdf

7. 临床研究SAS高级编程--SAS SQL
http://www.math.pku.edu.cn/teachers/lidf/course/gradsas/09.0%20SQL.pdf

8. Intermediate PROC SQL
http://www2.sas.com/proceedings/sugi23/Advtutor/p35.pdf

Friday, May 2, 2014

SAS forecast methods - Churn Management

1. Use SAS EG automatic churn management
http://www.analysisdatabase.com/descargas/ANALISIS%20DE%20ABANDONOS%20RETENCION.pdf


2.  SAS forecast methods
http://www.sascommunity.org/sugi/SUGI93/Sugi-93-67%20Little.pdf

http://www.iasri.res.in/sscnars/sas_manual/5-ts_sas_lecture.pdf
http://www.sascommunity.org/sugi/SUGI93/Sugi-93-67%20Little.pdf
http://www.nesug.org/Proceedings/nesug11/sa/sa10.pdf
http://support.sas.com/resources/papers/proceedings12/333-2012.pdf


SAS DateTime

http://analytics.ncsu.edu/sesug/2008/HOW-063.pdf

http://www.caloxy.com/papers/57-255-30.pdf

https://ciser.cornell.edu/sasdoc/saspdf/lrcon/c13.pdf