topic proc ucm forecast breakdown in SAS Forecasting and Econometrics https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/proc-ucm-forecast-breakdown/m-p/93880#M499 <HTML><HEAD></HEAD><BODY><P>Hi all,</P><P></P><P>I am kinda stuck at something and i would appreciate any help available.</P><P></P><P>I have ran a proc ucm for some sales data using a variable like promotion as well as level ,trend and season.</P><P></P><P>I proceed with a forecast to test my models fit.</P><P></P><P>In the outfor dataset i cant seem to be able to see the contribution in the forecasted value that is generated from my x variable (promotion), i only see from the unobserved components,</P><P></P><P>What could i do? cause my goal is to show in the end that the forecasted value consists of lets say 20% level, 20% season and a percent of my x variable(promotion).</P><P></P><P>Kind regards,</P></BODY></HTML> Wed, 27 Jun 2012 17:31:18 GMT chemicalab 2012-06-27T17:31:18Z proc ucm forecast breakdown https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/proc-ucm-forecast-breakdown/m-p/93880#M499 <HTML><HEAD></HEAD><BODY><P>Hi all,</P><P></P><P>I am kinda stuck at something and i would appreciate any help available.</P><P></P><P>I have ran a proc ucm for some sales data using a variable like promotion as well as level ,trend and season.</P><P></P><P>I proceed with a forecast to test my models fit.</P><P></P><P>In the outfor dataset i cant seem to be able to see the contribution in the forecasted value that is generated from my x variable (promotion), i only see from the unobserved components,</P><P></P><P>What could i do? cause my goal is to show in the end that the forecasted value consists of lets say 20% level, 20% season and a percent of my x variable(promotion).</P><P></P><P>Kind regards,</P></BODY></HTML> Wed, 27 Jun 2012 17:31:18 GMT https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/proc-ucm-forecast-breakdown/m-p/93880#M499 chemicalab 2012-06-27T17:31:18Z Re: proc ucm forecast breakdown https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/proc-ucm-forecast-breakdown/m-p/93881#M500 <HTML><HEAD></HEAD><BODY><P>Hi -</P><P>Not sure if this is what you were asking for, but check out the code below.</P><P>Thanks,</P><P>Udo</P><P></P><P>DATA WORK.airline;</P><P>&nbsp;&nbsp;&nbsp; LENGTH</P><P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Passengers&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 8</P><P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; DATE&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 8 ;</P><P>&nbsp;&nbsp;&nbsp; FORMAT</P><P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Passengers&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; BEST10.</P><P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; DATE&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; DATE9. ;</P><P>&nbsp;&nbsp;&nbsp; INFORMAT</P><P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Passengers&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; BEST10.</P><P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; DATE&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; DATE9. ;</P><P>&nbsp;&nbsp;&nbsp; INPUT</P><P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Passengers DATE;</P><P>if date="01SEP2001"d then event=1;</P><P>else event=0;</P><P>DATALINES;</P><P>37208190 01JAN1990</P><P>36078394 01FEB1990</P><P>44154535 01MAR1990</P><P>41355578 01APR1990</P><P>41234923 01MAY1990</P><P>44190605 01JUN1990</P><P>46036310 01JUL1990</P><P>48866148 01AUG1990</P><P>38989264 01SEP1990</P><P>41247525 01OCT1990</P><P>39346113 01NOV1990</P><P>39342008 01DEC1990</P><P>36716670 01JAN1991</P><P>33688621 01FEB1991</P><P>39477477 01MAR1991</P><P>39878137 01APR1991</P><P>41018839 01MAY1991</P><P>43166181 01JUN1991</P><P>45727485 01JUL1991</P><P>47802583 01AUG1991</P><P>38628284 01SEP1991</P><P>41170282 01OCT1991</P><P>37451867 01NOV1991</P><P>41159310 01DEC1991</P><P>36719242 01JAN1992</P><P>35757887 01FEB1992</P><P>41861145 01MAR1992</P><P>39651142 01APR1992</P><P>41800746 01MAY1992</P><P>47431207 01JUN1992</P><P>51806387 01JUL1992</P><P>53163782 01AUG1992</P><P>42510606 01SEP1992</P><P>42253840 01OCT1992</P><P>39566414 01NOV1992</P><P>40849861 01DEC1992</P><P>38529065 01JAN1993</P><P>36539132 01FEB1993</P><P>43775242 01MAR1993</P><P>43125070 01APR1993</P><P>44300992 01MAY1993</P><P>46334965 01JUN1993</P><P>49482654 01JUL1993</P><P>50934128 01AUG1993</P><P>43364388 01SEP1993</P><P>45930858 01OCT1993</P><P>42769377 01NOV1993</P><P>43872406 01DEC1993</P><P>40665800 01JAN1994</P><P>39264392 01FEB1994</P><P>48963885 01MAR1994</P><P>46124425 01APR1994</P><P>47883199 01MAY1994</P><P>50554571 01JUN1994</P><P>54026538 01JUL1994</P><P>54735918 01AUG1994</P><P>46918744 01SEP1994</P><P>49220347 01OCT1994</P><P>46742038 01NOV1994</P><P>47700305 01DEC1994</P><P>44687190 01JAN1995</P><P>41896823 01FEB1995</P><P>50985658 01MAR1995</P><P>49212746 01APR1995</P><P>49824307 01MAY1995</P><P>52881186 01JUN1995</P><P>54957670 01JUL1995</P><P>56585046 01AUG1995</P><P>47709592 01SEP1995</P><P>50595784 01OCT1995</P><P>48310520 01NOV1995</P><P>49158213 01DEC1995</P><P>45722827 01JAN1996</P><P>47130093 01FEB1996</P><P>55909079 01MAR1996</P><P>52390127 01APR1996</P><P>53777332 01MAY1996</P><P>56095234 01JUN1996</P><P>58058659 01JUL1996</P><P>59802408 01AUG1996</P><P>49758328 01SEP1996</P><P>53679750 01OCT1996</P><P>48464209 01NOV1996</P><P>53181172 01DEC1996</P><P>49228750 01JAN1997</P><P>47152265 01FEB1997</P><P>58163010 01MAR1997</P><P>53944329 01APR1997</P><P>55635847 01MAY1997</P><P>58172771 01JUN1997</P><P>61153800 01JUL1997</P><P>61907945 01AUG1997</P><P>51761004 01SEP1997</P><P>55026915 01OCT1997</P><P>51524528 01NOV1997</P><P>53801076 01DEC1997</P><P>49493435 01JAN1998</P><P>47935244 01FEB1998</P><P>58134392 01MAR1998</P><P>56871713 01APR1998</P><P>57787100 01MAY1998</P><P>60218576 01JUN1998</P><P>64890982 01JUL1998</P><P>62715323 01AUG1998</P><P>52507440 01SEP1998</P><P>56988546 01OCT1998</P><P>53904572 01NOV1998</P><P>55301435 01DEC1998</P><P>50996487 01JAN1999</P><P>49598613 01FEB1999</P><P>61104493 01MAR1999</P><P>58753327 01APR1999</P><P>59026650 01MAY1999</P><P>62607968 01JUN1999</P><P>66663065 01JUL1999</P><P>65309471 01AUG1999</P><P>55473871 01SEP1999</P><P>60425711 01OCT1999</P><P>58067248 01NOV1999</P><P>56156443 01DEC1999</P><P>51750091 01JAN2000</P><P>53589785 01FEB2000</P><P>64948291 01MAR2000</P><P>61998314 01APR2000</P><P>64349644 01MAY2000</P><P>67807620 01JUN2000</P><P>69852555 01JUL2000</P><P>68484270 01AUG2000</P><P>58107878 01SEP2000</P><P>62153577 01OCT2000</P><P>60213479 01NOV2000</P><P>57943798 01DEC2000</P><P>55108316 01JAN2001</P><P>52915977 01FEB2001</P><P>65068044 01MAR2001</P><P>62373365 01APR2001</P><P>62622321 01MAY2001</P><P>66193520 01JUN2001</P><P>69529176 01JUL2001</P><P>70133361 01AUG2001</P><P>38730506 01SEP2001</P><P>47746624 01OCT2001</P><P>48250267 01NOV2001</P><P>49816945 01DEC2001</P><P>47844272 01JAN2002</P><P>47055580 01FEB2002</P><P>59210469 01MAR2002</P><P>55059178 01APR2002</P><P>57283840 01MAY2002</P><P>60462592 01JUN2002</P><P>63263816 01JUL2002</P><P>63450360 01AUG2002</P><P>49890724 01SEP2002</P><P>58113746 01OCT2002</P><P>54351302 01NOV2002</P><P>60266707 01DEC2002</P><P>53399694 01JAN2003</P><P>50187136 01FEB2003</P><P>60504638 01MAR2003</P><P>56394469 01APR2003</P><P>58730501 01MAY2003</P><P>63367220 01JUN2003</P><P>68449976 01JUL2003</P><P>66720119 01AUG2003</P><P>54119709 01SEP2003</P><P>60225641 01OCT2003</P><P>56951134 01NOV2003</P><P>49866044.9 01DEC2003</P><P>44223734 01JAN2004</P><P>45343339 01FEB2004</P><P>;run;</P><P>proc ucm data = airline;</P><P>&nbsp;&nbsp; id date interval = month;</P><P>&nbsp;&nbsp; model passengers = event;</P><P>&nbsp;&nbsp; irregular;</P><P>&nbsp;&nbsp; level;</P><P>&nbsp;&nbsp; slope var = 0 noest;</P><P>&nbsp;&nbsp; season length = 12 type=trig;</P><P>&nbsp;&nbsp; estimate outest=est;</P><P>&nbsp;&nbsp; forecast back=12 lead=12 plot=(decomp forecasts) outfor=for;</P><P>run;</P><P>data _null_;</P><P> set est;</P><P> if upcase(component) = upcase("event") then do;</P><P>&nbsp; call symput('eventParm', estimate);</P><P>&nbsp; call symput('eventParmStd', std);</P><P> end;</P><P>run;</P><P>data adjust(keep=date event contrib stdContrib);</P><P> set airline(keep=date event);</P><P> est = symgetn('eventParm');</P><P>&nbsp;&nbsp;&nbsp; std = symgetn('eventParmStd');</P><P>&nbsp;&nbsp;&nbsp; contrib = est*event;</P><P>&nbsp;&nbsp;&nbsp; stdContrib = std*abs(event);</P><P>run;</P></BODY></HTML> Wed, 18 Jul 2012 18:52:03 GMT https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/proc-ucm-forecast-breakdown/m-p/93881#M500 udo_sas 2012-07-18T18:52:03Z