About RB1Kenobi

RB1Kenobi · ‎04-28-2015

Fair question RW9! I had thought of putting lists into a dataset but I had a nagging feeling that there might be a function to do the task that I wasn't aware of.

RB1Kenobi · ‎04-28-2015

Hi Tom I had to google "utility macro" and I've just glanced at this paper (http://www2.sas.com/proceedings/sugi30/029-30.pdf). I ran your example above but %Qlist was not recognized. So, is your macro %qlist something you've created as an autocall macro (could I see that code please?), or is it something that is there in the installation but not mapped on my machine? I tend to use enterprise guide by the way, version 5.1.

RB1Kenobi · ‎04-28-2015

Hello I'm using SAS9.3. Is anyone aware of a macro statement to remove quotes from a list, so that the same list could be used for multiple purposes? Example below where a list requires a quoted and unquoted version /* Using two macros - this is what I am trying to avoid. */ %let name_quoted='Mary' 'Philip'; %let name_unquoted =Mary Philip; proc report data=sashelp.class; title1 "Data for &name_unquoted."; /* list without quotes */ where name in (&name_quoted); /* list requires quotes */ run; All ideas are welcomed, however I was thinking there was a macro function that might work for this scenario? RB1

RB1Kenobi · ‎03-30-2015

Hello Tagsets is a good way to export to Excel. I find it quick to tweak what I know to suit different requirements (e.g. apply formatting). DDE is another technique but personally I found it a bit slow going as you have to define table ranges and locations specifically. Worth reading up the coding though as it may prove handy. I could not attach the file for some reason, so I've pasted an example below for you: /***** EXCEL OUTPUT *******/ ods tagsets.ExcelXP path= "###### type pathway here ##" /* Pathway ".....\...\..\.." */ file="two datasets.xls" /* filename wil XLS extention */ style=printer; ods escapechar="^"; /* Define the worksheet parameters */ ods tagsets.ExcelXP options(embedded_titles='yes' embedded_footnotes='yes' sheet_interval='none' /* Sheet interval determines whether to print on to the same sheet or a new one. */ sheet_name="Two datasets" zoom='80' skip_space='5,0,0,0,3' /* Determines the number of rows between the two tables. */ ); run; /* You can embed style elements within the proc print */ /* Dataset 1 */ proc print data= sashelp.class width=minimum label style (data) = [font_face=calibri just=left ] noobs; label age = "^{style[just=center fontweight=bold] AGE of Pupil}"; run;title;footnote; /* Dataset 2 */ proc print data= sashelp.fish (obs=100) width=minimum label style (data) = [font_face=calibri just=left] noobs; run;title;footnote; ods _all_ close; ods listing;

RB1Kenobi · ‎01-28-2015

Hello I'm running SAS 9.3 When I control the HAXIS by hard coding the character midpoints does this preclude me from using the AXIS1 options? In the example below I do not want a label on the horizontal axis but I can not fathom a method to use AXIS1 in conjunction with a character defined list. Axis1 STYLE=1 WIDTH=1 MINOR=NONE label=none ; PROC GPLOT DATA=sashelp.cars; PLOT invoice * origin/ VAXIS=AXIS2 HAXIS='Europe' 'Asia' 'USA' ; RUN; QUIT;

RB1Kenobi · ‎01-23-2015

I'm running 9.3 My ODS PDF gbarline chart is stretching the symbol so that a dot becomes a large elipse. The thing is, this is happening inconsistently as charts further within my program are appearing with the symbols shaped as expected. I'm struggling to identify & debug the culprit, suggestions please? Apologies - I can't replicate the issue with a demonstration program and I can't use my production data/chart in this forum. Thanks ods graphics on ; goptions device= jpeg border ; goptions htext=4 ftext="arial"; * Bar legend ; legend1 position=(middle left outside) origin=(,4.8)cm across=1 shape=bar(0.15,0.15)cm label=(position=(top ) j=l "VOLUME"); * Line plot legend ; legend2 position=(middle left outside) origin=(,1.8)cm across=1 repeat=2 label=(position=(top) j=l "VALUE") value=(height=3) ; axis1 value=(height=2.8) order=(&bar_order.) label=none; axis3 label=none ; symbol1 value=dot c=black h=4; symbol2 value=star c=black h=4; symbol3 value=square c=black h=4; symbol4 value=diamond c=black h=4; pattern1 c=biyg; pattern2 c=bio; * CHART - PDF; ods region width=10cm height=2cm x=1.2cm y=8cm; ods text = "^{style[font_face='calibri' fontsize=2.5 just=left fontweight=bold] demotext}"; ods region width=18cm height=7cm x=1.35cm y=8.5cm; proc gbarline data=work.a_dataset; title ' ' ; format vol val percent8.0 ; bar MONTH / discrete sumvar=VOLUME subgroup=PRODUCT legend=legend1 axis=axis3 maxis=axis1 clipref autoref missing ; plot / sumvar=Share_PRODUCT legend=legend2 axis=axis3 ; plot / sumvar=Share_SUB_PROD; run;

RB1Kenobi · ‎01-22-2015

Thanks Xia, that's really going to help keep my programs sleek.

RB1Kenobi · ‎01-22-2015

Hello I'm running SAS 9.3 In the example below, no data exists where product="B". But I'd like the proc report output to display "BRAVO, 0". I know I can undertake additional data-steps to force the existence of the required observations, but I wondered if there was a solution within PROC REPORT itself? proc format; value $ prodfull 'A'='Alpha' 'B'='Bravo' 'C'='Charlie' ; run; data demodata; do i=1 to 33; do t='A','C'; sales=i; product=t; output; sales=22;product="C"; output; end; end; run; Proc report data=work.demodata; title 'How can I display zero sales for Product Bravo?'; column Product sales; define product / group f=$prodfull. 'Product' ; define sales / sum 'Sales'; run;title;

RB1Kenobi · ‎01-21-2015

Using SYMBOL statements seems to work, although this example needs a bit of refinement. symbol2 c=CXDE7E6F interpol=box co=black; symbol1 c=CX7C95CA interpol=box co=black ;

RB1Kenobi · ‎01-21-2015

Hello I'm running SAS 9.03.01M2P081512. When I subgroup a bar on GBARLINE I can dictate the segment appearance using multiple PATTERN statements. However, in conjunction with multiple PLOT statements how can I align the pattern of the line with the relevant bar segment so that the legend is applicable to both the bar and the line? I'm happy to deviate from GBARLINE to achieve the desired results if required. The example below is amended from 33171 - Use PROC GBARLINE to create subgrouped bars. I would like the lines (PLOT) for Male and female height to match the legend for the bar. /* Set the graphics environment */ goptions reset=all cback=white border htitle=12pt htext=10pt; axis1 label=('Frequency of Age') minor=none; axis2 label=('Sum of Height') minor=none; symbol1 i=join c=black v=dot h=1.3; legend1 label=('Gender') value=('Female' 'Male') frame; pattern1 v=solid c=CXDE7E6F; pattern2 v=solid c=CX7C95CA; title1 'Using the SUBGROUP option with PROC GBARLINE'; data work.class_demo; set sashelp.class; if sex='M' then height_m=height; if sex='F' then height_f=height; run; /* Create the graph using the SUBGROUP= option */ /* on the BAR statement */ proc gbarline data= work.class_demo; bar age / subgroup=sex legend=legend1 discrete raxis=axis1 width=10; plot / sumvar= height_m raxis=axis2; plot / sumvar= height_f raxis=axis2 ; run; quit;

RB1Kenobi · ‎01-14-2015

I'm afraid I don't have an oracle background, but here's a suggestion by using a macro loop approach. * Define a macro to loop for your filters through; %macro tablexctract (tbl_filter=); * Open the Library and keep the table names that meet the "WHERE MEMNAME=" criteria; proc sql; create table Library_content as select memname as Dataset_Name , * from Dictionary.tables where LIBNAME='your libname here' and MEMTYPE='DATA' and memname like "%&tbl_filter.%" /** Additional "where" condition ***/ ;quit; /* Put the results into a macro variable. Caution - this approach relies on a single table name being returned from the filter above */ data _null_; set work.library_contentmaps (obs=1); call symputx('tblname',dataset_name); run; Proc sql; create table NEW_table_&tbl_filter. as select * from &tblname. ;quit; %mend; * Use the macro parameter TBL_FILTER to define the table name content you woudl like to extract; %tablexctract(tbl_filter=%str(emp); * loop 1; %tablexctract(tbl_filter=%str(...); * loop 2; %tablexctract(tbl_filter=%str(...); * loop 3; %tablexctract(tbl_filter=%str(...); * loop 4; %tablexctract(tbl_filter=%str(...); * loop 5; A very different approach would require you to research the autocall macro functionality, see %sysfunc examples 4 & 5 on this link: http://support.sas.com/documentation/cdl/en/mcrolref/62978/HTML/default/viewer.htm#p1o13d7wb2zfcnn19s5ssl2zdxvi.htm

RB1Kenobi · ‎01-14-2015

Hi I hope I've understood your question correctly. Option 1 - see the contents of a defined library: proc sql; create table Library_content as select memname as Dataset_Name , * from Dictionary.tables where LIBNAME='YOUR LIBNAME HERE' and MEMTYPE='DATA' ;quit; Option 2; Try Proc Contents to explore individual tables. Proc contents data=libname.tablename; run;

RB1Kenobi · ‎01-09-2015

Thanks Xia for your example. I'll try the ODS group. If it's not possible then that is good to know.

RB1Kenobi · ‎01-09-2015

Thanks RW9. However I'm basically trying to grow my proc report knowledge specifically to avoid the need to summarize the input data (less coding, perhaps less CPU, shorter program etc..). I was really looking to resolve the sort within the reporting procedure (if this is actually possible), it's just the ACROSS element that seems to prevent me ordering as I would like to.

RB1Kenobi · ‎01-09-2015

That's a useful paper, thanks. My attempts with the demo code (attached to the original post) are not working as expected. I think because the data contains a counter, so to apply a sort it would have to be summed prior to the report procedure? Would you be able to show it using my demo data please?

Online Status	Offline
Date Last Visited	‎05-26-2020 11:11 AM

Re: JSON MAP possibilities and limitations?

Re: JSON MAP possibilities and limitations?

Re: JSON MAP possibilities and limitations?

JSON MAP possibilities and limitations?

Re: ISO Date Format ("2018-09-04T14:11:58.000Z") into Date

Re: ISO Date Format ("2018-09-04T14:11:58.000Z") into Date

Re: ISO Date Format ("2018-09-04T14:11:58.000Z") into Date

Re: ISO Date Format ("2018-09-04T14:11:58.000Z") into Date

ISO Date Format ("2018-09-04T14:11:58.000Z") into Date

Re: Enterprise Guide - Slowly changing dimension?

Re: JSON MAP possibilities and limitations?

Re: ISO Date Format ("2018-09-04T14:11:58.000Z") into Date

Re: ISO Date Format ("2018-09-04T14:11:58.000Z") into Date

Re: Dynamic apply format by making the fmtname from a value in the sam...

Exporting to excel using tagsets excelxp and protect the file with pas...

Re: ODS PDF Proc Report Compute Block

Re: ODS PDF Proc Report Compute Block

Re: All_Tables or All_Columns

Re: Quote and Unquote a list?

Re: Quote and Unquote a list?

Quote and Unquote a list?

Re: how to export two datasets into single work sheet in excel

GPLOT HAXIS character defined limits options available?

Chart/ODS - why symbol is stretched shape?

Re: Proc Report - Print all format values even where no data?

Proc Report - Print all format values even where no data?

Re: GBARLINE - align pattern of segmented bar with plot?

GBARLINE - align pattern of segmented bar with plot?

Re: All_Tables or All_Columns

Re: All_Tables or All_Columns

Re: Proc Report - Sort descending the latest ACROSS column?

Re: Proc Report - Sort descending the latest ACROSS column?

Re: Proc Report - Sort descending the latest ACROSS column?