Data visualization with SAS programming

Blank space with proc template and sgrender

Reply
New Contributor RKC
New Contributor
Posts: 2

Blank space with proc template and sgrender

proc template;

define statgraph swimmer_doses_gtl;

begingraph / ;

/* creating an attribute map for future swimmer bars */

discreteattrmap name = 'bar_gtl';

value 'Head and neck' / fillattrs = (color= lightgreen) lineattrs = (color = black);

value 'Melanoma' / fillattrs = (color= cyan) lineattrs = (color = black);

value 'Bladder' / fillattrs = (color= purple) lineattrs = (color = black);

value 'Non-Small Cell Lung' / fillattrs = (color= gold) lineattrs = (color = black);

value 'Prostate' / fillattrs = (color= green) lineattrs = (color = black);

value 'Cervix' / fillattrs = (color= blue) lineattrs = (color = black);

value 'Malignant pleural mesothelioma (MPM)' / fillattrs = (color= pink) lineattrs = (color = black);

value 'Colon/Rectum' / fillattrs = (color= red) lineattrs = (color = black);

value 'Other, specify' / fillattrs = (color= brown) lineattrs = (color = black);

 

 

enddiscreteattrmap;

discreteattrvar attrvar = swimmer_doses_bar var = dose attrmap = 'bar_gtl';

 

/* creating an attribute map for future time-point responses */

discreteattrmap name = 'time_point_response_gtl';

value 'CR' / markerattrs = (color = black symbol = SquareFilled);

value 'PR' / markerattrs = (color = black symbol = CircleFilled);

value 'SD' / markerattrs = (color = black symbol = TriangleFilled);

value 'PD' / markerattrs = (color = black symbol = TriangleLeftFilled);

value 'NE' / markerattrs = (color = black symbol = StarFilled);

enddiscreteattrmap;

discreteattrvar attrvar = time_point_responses var = time_point_eval attrmap =

'time_point_response_gtl';

 

 

 

layout overlay

/

xaxisopts = (

type = linear

offsetmin =0.38 type=linear offsetmax = 0.08

label = 'Treatment Duration (days)'

labelattrs = (family = 'Arial' size = 10 pt)

linearopts = (tickvaluesequence = (start = 0 end = 250

increment = 10))

tickvalueattrs = (family = 'Arial' size = 10 pt)

)

yaxisopts = (

type = discrete

display = (label)

reverse = true

offsetmin = 0.125 offsetmax = 0.08

/*offsetmin = 0.125 offsetmax = 0.08*/

 

label = 'Subject (Treatment)'

labelattrs = (family = 'Arial' size = 10 pt)

)

/* second horizontal axis is needed for displaying patient ID */

x2axisopts = (

display = none

offsetmin = 0 offsetmax = 0

linearopts = (viewmin = 0 viewmax = 100)

)

;

/* drawing bars for duration of different dosages */

highlowplot

y = needed_order

low = dose_st_day

high = dose_en_day

/

type = bar

group = swimmer_doses_bar

outlineattrs = (pattern=solid)

barwidth = 0.50

name = 'dosages'

highcap=ongocap

;

/* drawing text information about subject ID */

scatterplot

x = id_subject_left

y = needed_order

/

xaxis = x2

datalabel = id_subject

datalabelposition = right

markerattrs = (size = 0)

datalabelattrs = (size = 9 pt color = black)

;

 

/* drawing markers to indicate time-point responses */

scatterplot

x = evald

y = needed_order

/

group = time_point_responses

name = 'responses'

markerattrs = (size = 6 pt);

 

 

/* Legend */

discretelegend 'responses'

/

location = inside

exclude = (' ')

halign = right

valign = bottom

border = false

down = 1

title = "Response:"

titleattrs = (size = 10 pt )

valueattrs = (size = 9 pt)

order=rowmajor

across=1

;

 

discretelegend 'dosages'

/

location = inside

exclude = (' ')

halign = right

valign = 0.1

border = false

down = 1

title = "Tumor Type:"

titleattrs = (size = 10 pt )

valueattrs = (size = 9 pt)

order=rowmajor

across=1

;

 

endlayout;

endgraph;

end;

run;

 

 

 

Hi ,

Too much space creating at bottom and left side of swimmer plot and further when I give footnote graph is shrinking and text overlapping. Any help is appreciated. Thanks.

 

SAS Super FREQ
Posts: 1,265

Re: Blank space with proc template and sgrender

The blank space you have around the graph is outside the definition of the graph.  Check your ODS RTF destination or any ODS LAYOUT settings.  Also, I suggest you could use the LOWLABEL of the HighLow plot itself for the subject id.  Or, an AxisTable.

Ask a Question
Discussion stats
  • 1 reply
  • 142 views
  • 1 like
  • 2 in conversation