I've got some code that assigns a unique ID to observations in a SAS table. This unique ID is then used to generate graphs using SGPLOT corresponding to that ID, and insert them as .PNG files into an email.
This code runs fine in SAS 7.11 but I'm migrating the existing code to SAS 8.3, and SAS 8.3 appears to behave differently.
- First, the 'volume' observations get an image ID equal to ((_n_ * 4)-3) and the 'percentage' observations get an image ID equal to ((_n_ * 4)-1).
id_num=_N_; /* Images generate skipping every second number, the formulas below align the numbers to the charts so we know whats what when building the email. */
Volume_ImageID = ((_n_ * 4)-3);
Percent_ImageID = ((_n_ * 4)-1);
Volume_text=compress(put(Volume_ImageID,best3.));
Percent_text=compress(put(Percent_ImageID,best3.));
Volume_attach_text=catt('"/email',compress(put(Volume_ImageID,best3.)),'.png"'," inlined='image",compress(put(Volume_ImageID,best3.)),"'");
Percent_attach_text=catt('"/email',compress(put(Percent_ImageID,best3.)),'.png"'," inlined='image",compress(put(Percent_ImageID,best3.)),"'");
2. The dataset generated from above gives me the following, which is correct:
3. I then have a macro which outputs the generated graphs for each unique ID. The full macro code is further down below.
ods listing gpath="&workdir.";
ods graphics on / /*reset=index(1)*/ outputfmt=png imagename='email';4. In SAS 7.11, the code above somehow spits out the following .png files, which is correct:
email1.png
email3.png
and so on, based on the number of loops/observations.
5. In SAS 8.3, it spits out:
email.png
email1.png
which generates a file not found error when the email macro tries to insert email3.png or any additional .PNGs.
6. I've looked around resetting indexes (starting the index from 1 doesn't solve the fact that the files named using increments of 2), using BY values, but I'm a bit over my head as the code was written by a former employee. I'm hoping there's an easier way of assigning an ID to the observations which will also make outputting the .PNG files easier.
Any help would be greatly appreciated, please let me know if you need more sample code or data.
Full macro code from Point 3, which generates the charts using SGPLOT and outputs them as .PNG files.
%macro generate_charts(sel_QMBL, sel_category);
%let max_total_count=0;
%let max_queue_est_count=0;
%let max_outer_high=0;
%let vol_max_count=0;
DATA VRNTCLK.CC_DATA_TEST;
SET VADROP.UC9_SA_VERINT_CC;
Category=TRANWRD(category,'z_','');
Category=TRANWRD(category,'T_','');
where (TRANWRD(category,'z_','')=&&sel_category or TRANWRD(category,'T_','')=&&sel_category)
and MBL = &&sel_QMBL
and segment='Total MBL Cat'
and ddate between &st_date and &_repdate3;
RUN;
proc sort data=VRNTCLK.CC_DATA_TEST;
by ddate MBL;
run;
/* max chart values */
PROC SQL noprint;
select max(total_count) as v1,
max(queue_est_count) as v2,
max(outer_high) as v3,
round(max((calculated v1), (calculated v2), (calculated v3)) + 4,10) as v4
into :max_total_count,
:max_queue_est_count,
:max_outer_high,
:vol_max_count
from VRNTCLK.CC_DATA_TEST;
QUIT;
ods listing gpath="&workdir.";
ods graphics on / /*reset=index(1)*/ outputfmt=png imagename='email';
title &sel_category ' on ' &sel_QMBL;
proc sgplot data=VRNTCLK.CC_DATA_TEST noautolegend noborder pad=0;
/* title1 Volume Control Chart for &_repdate2;
title2 &sel_category ' on ' &sel_QMBL; */
styleattrs datasymbols=(circlefilled)
datalinepatterns=(solid);
band x=ddate lower=outer_low upper=outer_high /fillattrs=(color='#c9f1ff');
band x=ddate lower=middle_low upper=middle_high /fillattrs=(color='#82d0f8');
band x=ddate lower=inner_low upper=inner_high /fillattrs=(color='#00b0f0');
series y=Day20MA x=ddate /group=category lineattrs=(Color='#9b0503' thickness=3 pattern=longdash);
series y=Queue_Est_Count x=ddate /group=category lineattrs=(color='#3d5aae' thickness=3);
format ddate date9.;
xaxis min=&st_date max=&_repdate3 offsetmin=0 offsetmax=0 display=(nolabel) type=time interval=day /* grid */;
yaxis display=(nolabel) min=0 /*values=(0 to &vol_max_count)*/ /* grid */;
INSET "&&sel_category. on &&sel_QMBL." / POSITION = TOPRIGHT BORDER;
run;
proc sgplot data=VRNTCLK.CC_DATA_TEST noautolegend noborder pad=0;
/* title1 Percentage Control Chart for &_repdate2;
title2 &sel_category ' on ' &sel_QMBL; */
styleattrs datasymbols=(circlefilled)
datalinepatterns=(solid);
band x=ddate lower=outer_low_perc upper=outer_high_perc /fillattrs=(color='#c9f1ff');
band x=ddate lower=middle_low_perc upper=middle_high_perc /fillattrs=(color='#82d0f8');
band x=ddate lower=inner_low_perc upper=inner_high_perc /fillattrs=(color='#00b0f0');
series y=Day20MA_perc x=ddate /group=category lineattrs=(Color='#9b0503' thickness=3 pattern=longdash);
series y=perc_rcvd x=ddate /group=category lineattrs=(color='#3d5aae' thickness=3);
format ddate date9.;
xaxis min=&st_date max=&_repdate3 offsetmin=0 offsetmax=0 display=(nolabel) type=time interval=day /* grid */;
yaxis display=(nolabel) min=0 /* grid */;
format Day20MA_Perc percent8.4;
Format Day20StdDev_Perc percent8.4;
Format Inner_High_Perc percent8.4;
Format Inner_Low_Perc percent8.4;
Format Middle_High_Perc percent8.4;
Format Middle_Low_Perc percent8.4;
Format Outer_High_Perc percent8.4;
Format Outer_Low_Perc percent8.4;
Format Perc_Rcvd percent8.4;
Format queue_ans_perc percent8.4;
INSET "&&sel_category. on &&sel_QMBL." / POSITION = TOPRIGHT BORDER;
run;
%mend generate_charts;
Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.
-
1959 views
-
0 likes
-
-
