Statystyka od Podstaw z wykorzystaniem narzędzi SAS

Nr przedmiotu: 136030-0131 SLLD

Prowadzący: dr hab. Ewa Frątczak, dr Sebastian Zając
Zakład Analizy Historii Zdarzeń i Analiz Wielopoziomowych

Ćwiczenia 2

Libname sop 'C:\sciezka do katalogu';

%LET nazwa_zmiennej = wartosc;
&nazwa_zmiennej

proc contents data=sop.diagnoza07;
run;

proc sql;
create table sop._diag07 as
select * from sop.diagnoza07 t1
where warunek
quit;

proc sql;
create table sop.diag_max as
select t1.woj, (max(t1.dochm)) LABEL="max dochodow" as max
from sop.diagnoza07 t1
group by t1.woj;
quit;

%LET file = "C:\katalog\output.rtf"
ods rtf file=&file;
proc print data=sop.diag_max;
var woj suma max;
run;
ods rtf close;

proc surveyselect data=sop.diagnoza07
out=sop.sample
method=srs
n=50;
run;
proc sql;
...
quit;

proc gchart data=sop.graph1;
vbar woj / type=pct discrete outside=percent;
run;

proc sql;
create table sop.graph1 as
select t1.woj
,(Mean(t1.dochm)) label="mean" format=nlmnlpln12.2 as mean
from sop.diagnoza07 t1
group by t1.woj;
quit;

proc gchart data=sop.graph1;
vbar woj / sumvar=mean frame
discrete descending
coutline=black;
format mean nlmnipln12.2;
run;

symbol value=dot c=orange i=none;
proc gplot data=sop.diagnoza07;
plot dochm*edu;
run;

proc gchart data=sop._diag07_max;
pie woj / sumvar=suma discrete slice=arrow;
where woj in (2 4);
run;

data kursy;
input t ddmmyy6. USD EUR EXPORT;
format t poldfmn.;
cards;
010113 3.1145 4.1369 16.25
010213 3.1197 4.1734 16.91
010313 3.2058 4.1581 15.87
010413 3.1825 4.1409 16.76
010513 3.2149 4.1759 15.14
010613 3.2512 4.2865 16.77
010713 3.2688 4.2756 16.41
010813 3.1767 4.2295 .
010913 3.1750 4.2376 .
011013 3.0704 4.1908 .
;
run;

goptions reset=all i=join;
axis1 order=10 to 20;
axis2 label=("PLN");
legend1 label=none;
proc gplot data=kursy;
plot USD * t EUR * t / overlay vaxis=axis2 legend=legend1;
plot2 EXPORT * t / vaxis=axis1 legend=legend1;
run;

proc freq data=sop.diagnoza07;
tables woj / plots=freqplot;
run;

proc means data=sop.diagnoza07
mean std stderr var min max mode range sum n nmiss q1 median q3;
var dochm;
run;

proc univariate data=sop.diagnoza07 noprint;
var dochm;
histogram;
run;

proc sqplot data=sop.diagnoza07;
vbox dochm;
run;

data sop.d;
set sop.diagnoza07(where=(dochm>0));
plectest=(plec>1);
run;
proc means data=sop.d vardef=df
mean std stderr var min max mode range sum n nmiss q1 median q3;
var dochm;
class plectest;
run;

proc sqplot data=sop.d;
vbox dochm / group=plectest;
run;
proc sgpanel data=sop.d;
panelby plectest / columns=1;
histogram dochm;
run;

data s;
do i=1 to 10;
y=ranbin(20151006,1,0.5);
output;
end;
run;
goptions reset=all;
symbol interpol=needle v=dot;
axis1 order=(0 to 10 by 1);
proc gplot data=s;
plot y * i = y / haxis=axis1;
run;

data normal;
do x=1 to 100 by 1;
wzrost=int(rannor(20151006)*22+177);
output;
end;
run;
proc sgplot data=normal;
histogram wzrost;
run;
proc means data=normal n mean maxdec=0;
var wzrost;
run;

data s;
do x=0 to 10;
y_pdf = pdf('binomial',x,0.3,10);
y_cdf = cdf('binomial',x,0.3,10);
output;
end;
run;
data s;
set s;
cumulative + y_pdf;
run;
symbol1 interpol=needle v=dot color=blue;
symbol2 interpol=needle v=dot color=red;
axis1 order=(0 to 10 by 1);
axis2 order=(0 to 1 by 0.1) label=("P(X = x)");
proc gplot data=s;
plot y_pdf * x / haxis=axis1 vaxis=axis2;
plot2 y_cdf * x /vaxis=axis2;
run;