data ppp;
  input mese$ anno NI;
cards;
01 2016 99.6
02 2016 99.4
03 2016 99.6
04 2016 99.5
05 2016 99.8
06 2016 99.9
07 2016 100.1
08 2016 100.3
09 2016 100.1
10 2016 100
11 2016 99.9
12 2016 100.3
01 2017 100.6
02 2017 101
03 2017 101
04 2017 101.4
05 2017 101.2
06 2017 101.1
07 2017 101.2
08 2017 101.5
09 2017 101.2
10 2017 101
11 2017 100.8
12 2017 101.2
01 2018 101.5
02 2018 101.5
03 2018 101.8
04 2018 101.9
05 2018 102.2
06 2018 102.4
07 2018 102.7
08 2018 103.1
09 2018 102.6
10 2018 102.6
11 2018 102.4
12 2018 102.3
01 2019 102.4
02 2019 102.5
03 2019 102.8
04 2019 103
05 2019 103
06 2019 103.1
07 2019 103.1
08 2019 103.5
09 2019 102.9
10 2019 102.8
11 2019 102.6
12 2019 102.8
01 2020 102.9
02 2020 102.8
03 2020 102.9
04 2020 103
05 2020 102.8
06 2020 102.9
07 2020 102.7
08 2020 103
09 2020 102.3
10 2020 102.5
11 2020 102.4
;
run;
data date_tutte;
	format data_tutte ddmmyy10.;
	set ppp;
	anno1=trim(left(put(anno,z4.)));
 	mese1=trim(left(mese));
	data_=(trim(left(anno1))||trim(left(mese1))||trim(left("01")));
	data_tutte=input(data_,yymmdd8. );
	keep NI data_tutte;
run;

proc sgplot data=date_tutte ;
	series x=data_tutte y=NI;	
run;

PROC TIMESERIES DATA=date_tutte OUT=date_tutte_s SEASONALITY=12 OUTDECOMP=deseason plot=(series 
		 decomp SA);
	VAR NI;
	DECOMP /MODE=MULT;
quit; run;

proc sgplot data=deseason ;
	series x=_TIMEID_ y=ORIGINAL;	 
	series x=_TIMEID_ y=SA ;	 
run;