@arthurdpereira , entendo sua preocupação
Espero que ajude, tenho um exemplo carregando um shapefile com dados do México, talvez te ajude como referência de como fazer.
/************************************************
SAS Visual Analytics Custom Polygon Deployment Program
Use this program to import custom polygons from an Esri SHP file for use in SAS Visual Analytics.
Before running, define the two Librefs and the block of macro variables at the top of this program,
and then modify Step 2 if needed to ensure that the values of the IDLABEL (IDNAME) column are unique.
For details, see the SAS Visual Analytics 7.4: Administration Guide at support.sas.com.
************************************************/
/* ====== Update the following code: ===== */
libname MAPSCSTM "D:\opt\sasinside\SASHome\SASFoundation\9.4\mapscstm"; /* The library containing the custom polygon data (For example, C:\Data\mapscstm) */
libname VALIB "D:\opt\sasinside\Config\Lev1\SASApp\Data\valib"; /* The library containing the Visual Analytics lookup tables (For example, C:\SAS\VA\Lev1\SASApp\Data\valib) */
%let REGION_LABEL=Mexico States from Shapefile; /* The label for the custom region (For example, My Sales Regions) */
%let REGION_PREFIX=M1; /* unique ISO 2-Letter Code - make sure it doesn't conflict with any code! (For example, CS). You can run a query in attrlookup to check availability */
%let REGION_ISO=000; /* unique ISO Code - make sure it doesn't conflict with any other code! (For example, 000). You can run a query in attrlookup to check availability */
%let REGION_DATASET=MAPSCSTM.MX_STATES_SHP1; /* Name of the polygon data set to be created - be sure to use suffix "1" (For example, MAPSCSTM.MYSALESREGION1) */
%let REGION_SHP_FILE=D:\Mexico_States\Mexico_States.shp; /* Source polygon file in Esri SHP format (For example, C:\Data\myregion.shp) */
%let REGION_SHP_ID_COL=CODE; /* the column representing the unique ID of a polygon (For example, ID). You can find it after running PROC MAPIMPORT below */
%let REGION_SHP_NAME_COL=NAME; /* the column representing the name of the polygon (For example, NAME). You can find it after running PROC MAPIMPORT below */
%let REGION_SHP_DENSITY=7; /* Value between 1 (low) - 10 (high) describing the detail level/density. (For example, 3) */
/* ====== End code updates -- Also see the section marked as Step 2 below ===== */
/* Utility macro to stop processing in case of any validation errors */
%let ERROR_COUNT=0; /* 0 indicates no error */
%let ERROR_MESSAGE=;
%MACRO ErrorTest;
%GLOBAL ERROR_COUNT;
%IF &ERROR_COUNT gt 0 %THEN
%DO;
%PUT ERROR: &ERROR_MESSAGE;
%abort cancel;
%let ERROR_COUNT=0; /* reset */
%END;
%MEND;
/* [BACKUP] Create a backup of the original lookup tables if required */
%macro backupLookupTable(dsn);
%if %sysfunc(exist(&dsn._original)) %then
%do;
%put INFO: Backup table &dsn found.;
%end;
%else
%do;
data &dsn._original;
set &dsn;
run;
%end;
%mend backupLookupTable;
%backupLookupTable(valib.attrlookup);
%backupLookupTable(valib.centlookup);
/* [VALIDATION STEP] Check for unique 2-letter ISO code */
proc sql noprint;
/* count and create error code if greater than zero */
select count(ID) as count,
case
when calculated count > 0 then "2-letter ISO code ®ION_PREFIX. is not unique in ATTRLOOKUP. Please specify a different code."
else ""
end
into :ERROR_COUNT, :ERROR_MESSAGE
from valib.attrlookup_original
where upcase(ID) eq upcase("®ION_PREFIX.");
quit;
%ErrorTest;
/* [VALIDATION STEP] Check for unique ISO code */
proc sql noprint;
/* count and create error code if greater than zero */
select count(ISO) as count,
case
when calculated count > 0 then "ISO code ®ION_ISO. is not unique in ATTRLOOKUP. Please specify a different ISO code."
else ""
end
into :ERROR_COUNT, :ERROR_MESSAGE
from valib.attrlookup_original
where ISO eq "®ION_ISO.";
quit;
%ErrorTest;
/* Remove any previous entries if required */
proc sql;
delete from valib.attrlookup where ID like "®ION_PREFIX.%";
delete from valib.centlookup where ID like "®ION_PREFIX.%";
quit;
/* Step 1: Import the SHP file into SAS */
PROC MAPIMPORT DATAFILE="®ION_SHP_FILE."
OUT=®ION_DATASET.;
ID ®ION_SHP_ID_COL.;
RUN;
/* Add the DENSITY column to the polygon data set used to reduce the polygon density in the visualization */
proc greduce data=®ION_DATASET. out=®ION_DATASET.;
id ®ION_SHP_ID_COL.;
run;
/* Determine the column type of REGION_SHP_ID_COL and REGION_SHP_NAME_COL */
data _null_;
set ®ION_DATASET.(obs=1);
call symputx('REGION_SHP_ID_COL_TYPE', vtype(®ION_SHP_ID_COL.));
call symputx('REGION_SHP_NAME_COL_TYPE', vtype(®ION_SHP_NAME_COL.));
run;
%put &=REGION_SHP_ID_COL_TYPE;
%put &=REGION_SHP_NAME_COL_TYPE;
/* Step 2: Region lookup can either be via IDLABEL (IDNAME) or ID. In case we lookup by IDLABEL let's generate a
unique name. The following steps might need to be adjusted dependent on your source data.
Depending on your source data, you might want to concatenate the ®ION_SHP_NAME column with another column
in order to ensure that the values are unique.
*/
/*===== Update the following code as needed -- see the comment block above =====*/
%macro createIdFormat();
proc sql;
create table work.idname_fmt as
select distinct
"idname" as FMTNAME,
"®ION_SHP_ID_COL_TYPE." as TYPE,
®ION_SHP_ID_COL. as START,
%if %upcase(®ION_SHP_NAME_COL_TYPE.) eq N %then %do;
/* convert column to character if required */
compress(put(®ION_SHP_NAME_COL.,best32.))
%end;
%else %do;
trim(left(®ION_SHP_NAME_COL.))
%end;
as LABEL
from ®ION_DATASET.;
quit;
%mend createIdFormat;
%createIdFormat;
/*===== End code updates =====*/
/* [VALIDATION STEP] Check for duplicate names in the IDNAME lookup data set. With duplicate entries here the idname based lookup would not work.
If you have names listed here, revisit the previous step to make sure each LABEL is unique, for example, by merging other parent level details */
proc sql noprint;
create table work.nonunique_names
as select distinct LABEL, count(LABEL) as count
from work.idname_fmt
group by LABEL
having calculated count > 1
order by count desc;
/* count and create error code if greater than zero */
select count(*) as count,
case
when calculated count > 0 then compress(put(calculated count,8.))
|| " duplicate names have been found. Please check the idname_fmt table for duplicates."
|| " A list of duplicate names are in WORK.NONUNIQUE_NAMES."
else ""
end
into :ERROR_COUNT, :ERROR_MESSAGE
from work.nonunique_names;
quit;
%ErrorTest;
proc format cntlin=work.idname_fmt;
run;
/* [VALIDATION STEP] Check for required columns in the input polygon data set. */
data _null_;
dsid = open("®ION_DATASET.");
array colNames(5) $50 _temporary_ ("®ION_SHP_ID_COL.", "X", "Y", "SEGMENT", "DENSITY");
do i = lbound(colNames) to hbound(colNames);
check = varnum(dsid, colNames(i));
if check eq 0 then
do;
call symputx('ERROR_COUNT', 1);
call symputx('ERROR_MESSAGE',"Column " || compress(colNames(i)) || " is required in ®ION_DATASET.");
stop;
end;
end;
run;
%ErrorTest;
/* [VALIDATION STEP] Check for conflicting columns in the input polygon data set. */
data _null_;
array colNames(7) $50 _temporary_ ("IDNAME", "LONG", "LAT", "RESOLUTION", "LAKE", "ISOALPHA2", "AdminType");
do i = lbound(colNames) to hbound(colNames);
if colNames(i) eq "®ION_SHP_ID_COL." then
do;
call symputx('ERROR_COUNT', 1);
call symputx('ERROR_MESSAGE',"Column name for REGION_SHP_ID_COL conflicts with predefined column " || compress(colNames(i)) || ".");
stop;
end;
end;
run;
%ErrorTest;
/* Step 3: Create the final polygon data set with the required set of columns. We are going to create a new
shorter ID column here as the original ID columns are often long and unnecessarily complex. Note, we also apply
the previously created idname format to the IDNAME column.
*/
data ®ION_DATASET.;
set ®ION_DATASET.(keep=®ION_SHP_ID_COL. X Y SEGMENT DENSITY);
/* create new (shorter) ID */
length _TMP_ $60. _SID_ 8.;
if (_n_ eq 1) then
do;
_SID_ = 1;
_TMP_ = ®ION_SHP_ID_COL.;
end;
if _TMP_ ne ®ION_SHP_ID_COL. then
_SID_ = _SID_ + 1;
_TMP_ = ®ION_SHP_ID_COL.;
retain _TMP_ _SID_;
drop _TMP_ _SID_;
_RID_ = compress("®ION_PREFIX.-" || put(_SID_,8.));
IDNAME = put(®ION_SHP_ID_COL., idname.);
LONG = X;
LAT = Y;
ISO = "®ION_ISO.";
RESOLUTION = 1;
LAKE = 0;
ISOALPHA2 = "®ION_PREFIX.";
AdminType = "regions";
where density<®ION_SHP_DENSITY.;
keep _RID_ SEGMENT IDNAME LONG LAT X Y ISO DENSITY RESOLUTION LAKE ISOALPHA2 AdminType;
rename _RID_=ID;
run;
/* Step 4: Add a custom region (country) to ATTRLOOKUP - a new entry will appear in the VA country selection list */
proc sql;
insert into valib.attrlookup
values (
"®ION_LABEL.", /* IDLABEL=State/Province Label */
"®ION_PREFIX.", /* ID=SAS Map ID Value */
"®ION_LABEL.", /* IDNAME=State/Province Name */
"", /* ID1NAME=Country Name */
"", /* ID2NAME */
"®ION_ISO.", /* ISO=Country ISO Numeric Code */
"®ION_LABEL.", /* ISONAME */
"®ION_LABEL.", /* KEY */
"", /* ID1=Country ISO 2-Letter Code */
"", /* ID2 */
"", /* ID3 */
"", /* ID3NAME */
0 /* LEVEL (0=country level, 1=state level) */
);
quit;
/* Register each single region in the lookup data set */
proc sql;
insert into valib.attrlookup
select distinct
IDNAME, /* IDLABEL=State/Province Label */
ID, /* ID=SAS Map ID Value */
IDNAME, /* IDNAME=State/Province Name */
"®ION_LABEL.", /* ID1NAME=Country Name */
"", /* ID2NAME */
"®ION_ISO.", /* ISO=Country ISO Numeric Code */
"®ION_LABEL.", /* ISONAME */
trim(IDNAME) || "|®ION_LABEL.", /* KEY */
"®ION_PREFIX.", /* ID1=Country ISO 2-Letter Code */
"", /* ID2 */
"", /* ID3 */
"", /* ID3NAME */
1 /* LEVEL (1=state level) */
from ®ION_DATASET.;
quit;
/* Register custom region in CENTLOOKUP */
proc sql;
/* Add custom region */
insert into valib.centlookup
select distinct
"®ION_DATASET." as mapname,
"®ION_PREFIX." as ID,
avg(x) as x,
avg(y) as y
from ®ION_DATASET.;
/* Add custom provinces */
insert into valib.centlookup
select distinct
"®ION_DATASET." as mapname,
ID as ID,
avg(x) as x,
avg(y) as y
from ®ION_DATASET.
group by id;
quit;
/* Step 5: (optional) Creates a validation data set which can be used to validate the newly added regions in SAS Visual Analytics */
proc sql;
create table ®ION_DATASET._VALIDATE as
select distinct ID as ID,
IDNAME as NAME
from ®ION_DATASET.;
create table ®ION_DATASET._VALIDATE as
select *,
round(ranuni(1) * 10000) as measure1,
round(ranuni(1) * 100000) as measure2 format=dollar20.0
from ®ION_DATASET._VALIDATE
group by ID, NAME
order by ID, NAME;
quit;
... View more
