#statistical graphics

[rgraphs]

[prachivermablr]

16 Important SAS/STAT Features You Must Know - DataFlair

What is SAS/STAT features,benefits of STAT Software of SAS,Analysis of Variance,Bayesian Analysis,Multivariate Analysis,Regression, Standardization

[rgraphs]

Size-Ordered Comparison Bars with Labels

%macro doit(var,name); proc sql; create table &name.overall as select ver, sum(&var) as &var from check group by ver ; create table &name.state as select ver, d_state as state, sum(&var) as &var from check where state in(&statelist) group by ver, state ; create table &name.statelabs as select state, sum(&var*(ver='Alternative')) as num, sum(&var*(ver='Standard ')) as den, (calculated num-calculated den)/(calculated den) as lab format=percent8.1, 'Alternative' as ver, max(calculated num,calculated den) as ypos from &name.state group by state ; quit; %mend; %doit(q1b,q1) %doit(q2e,q2) %doit(q3i,q3) %doit(q4,q4) %macro plotit(var,name,lab); &name.sub<-subset(check,&var._f==4,); p1<-ggplot(&name.sub,aes(x=&var,color=ver))+ stat_ecdf()+ scale_x_log10(labels=comma)+ xlab("\n&lab")+ ylab("Cumulative Density\n")+ guides( color=guide_legend("Imputation Version:",keyheight=1,keywidth=2) ); CairoPNG("&name.ECDF.png", width=6, height=6, units="in", res=500); print(p1); dev.off(); &name.state<-transform(&name.state,state=reorder(state,&var)); p2<-ggplot(&name.state,aes(x=state,y=&var,fill=ver))+ geom_bar(position="dodge",stat="identity",color="black")+ geom_text(data=&name.statelabs,aes(x=state,y=ypos+10000,label=lab))+ scale_y_continuous(labels=comma)+ xlab("\nState")+ ylab("&lab\n")+ guides( fill=guide_legend("Imputation Version:",keyheight=1,keywidth=2) ); CairoPNG("&name.StateBars.png", width=10, height=5, units="in", res=500); print(p2); dev.off(); %mend; data _null_; %RCodeOpen( datforR=check q1state q1statelabs q2state q2statelabs q3state q3statelabs q4state q4statelabs meths, stripformats=0, menable=1 ) datalines; library(ggplot2) library(scales) library(gridExtra) library(Cairo) t<-theme_set(theme_bw()) t<-theme_update( panel.grid.minor.x=element_blank(), panel.grid.minor.y=element_blank(), panel.background=element_blank(), axis.text.x=element_text(angle=45, hjust=1, vjust=1), strip.background=element_rect(fill="white"), legend.position="bottom", legend.direction="horizontal", legend.justification=c(0,0), legend.box.just="left", plot.title=element_text(size = rel(1.2)) ) setwd("/rtpnfil03/rtpnfil03_vol4/ASPP/Imputation_Weighting/Programs/2016/impccompare") %plotit(q1b,q1,January 1 Population) %plotit(q2e,q2,Entries) %plotit(q3i,q3,Exits) %plotit(q4,q4,December 31 Population) pmeth<-ggplot(meths,aes(x=D_State,fill=entmeth))+ geom_bar(color="black",position="dodge")+ scale_y_continuous(breaks=seq(0,16,by=2))+ xlab("\nState")+ ylab("Method Frequency\n")+ guides( fill=guide_legend("Standard Imp Approach:",keyheight=1,keywidth=2) ) CairoPNG("Methods.png", width=10, height=5, units="in", res=500) print(pmeth) dev.off() ; %RCodeClose

[rbusinessanalytics]

R Programming: The Past

This blog introduces you to the world of ‘R’. R is just not a word rather R is marvelous language and an amazing tool used for Data analysis and Statistical Graphics. It adds an easy visualization to your data, it is free, and widely used by professional statisticians and market analysts. So, before starting with something amazing you should be aware about its past. In this blog you’ll get to know how R has developed. R has transformed from S language and S language itself had few phases in its whole existence. Regardless S is not with us now, still you should know a bit about the versions of S language as R is evolved from S. S VERSION 1 S VERSION 2 S VERSION 3 S VERSION 4 •Implementation nearly all FORTRAN based, via preprocessing tools • Only for our (bizarre) operating system • Adopted our existing graphics & data structure software • Portability via a Unix implementation • Most features carried over from V. 1 • Device-independent graphics • Merged some new ideas with S • “Everything is an object” (including functions) • No direct back compatibility with S2 • ‘Computing with data’ distinguished from statistical computing • Extensions to the S programming model • Today we have the S language, implemented in R HISTORY OF • The R language is a terminology of S language which was designed in the 1980s by John Chambers at Bell labs and has been in widespread use in the statistical community since. • It was grown up by Robert Gentleman and Ross Ihaka of the University of Auckland. R has been with us since 1993. • After the four versions of ‘S’ language we came up with R tool and R language. • During a chance they both encounter in the corridor, the following conversation took place: Gentleman: “Let’s write some software.” Ihaka: “Sure, that sounds like fun.” • It is concluded to adopt the syntax of the S language which has developed at Bell Laboratories. – As a joke, the name “R” is punched for the language (putting for Robert and Ross). LICENSING When an initial version of R language has completed then colleagues of Robert and Ross encouraged them to release the language as “free software”. Finally colleagues made them convinced and they adopted the Free Software Foundation GPL as their license and began to make releases via Internet. They started a small email list so that they and the users can discuss R. So this was all about history and past of R language and in my next blog I will be discussing the present status of R.

[sunfoundation]