Files
SAS_PACKAGES/packages/macroarray.md
yabwon b373917770 SAS Packages Framework, version 20201010
**SAS Packages Framework**, version 20201010:
- Improvement in testing facility for the framework.
- Change in SAS components testing, missing component issues a *warning* instead of *error*.
- Documentation updated, `SPFinit.md` file added.
- Minor bug fixes.

Packages recompiled with new version of SAS Packages Framework:
- `SQLinDS` (version 2.2)
- `macroArray` (version 0.5)
- `DFA` (version 0.2)
- `BasePlus` (version 0.7)
  - documentation updated
  - new macro `%symdelGlobal()` added
- `dynMacroArray` (version 0.2)
2020-10-10 19:26:49 +02:00

1169 lines
36 KiB
Markdown

- [The macroArray package [ver. 0.5]](#macroarray)
- [Content description](#content-description)
* [`%appendArray()` macro](#appendarray-macro)
* [`%appendCell()` macro](#appendcell-macro)
* [`%array()` macro](#array-macro)
* [`%concatArrays()` macro](#concatarrays-macro)
* [`%deleteMacArray()` macro](#deletemacarray-macro)
* [`%do_over()` macro](#do-over-macro)
* [`%do_over2()` macro](#do-over2-macro)
* [`%do_over3()` macro](#do-over3-macro)
* [`%make_do_over()` macro](#make-do-over-macro)
* [License](#license)
---
# The macroArray package [ver. 0.5] <a name="macroarray-package"></a> ###############################################
The **macroArray** package implements a macro array facility:
- `%array()`,
- `%do_over()`,
- `%make_do_over()`,
- `%deletemacarray()`,
- `%concatarrays()`,
- `%appendcell()`.
- etc.
The set of macros, which emulates classic
data-step-array functionality on the macro
programming level, is provided.
*Note:*
If you are working with BIG macroarrays do not
forget to verify your session setting for macro
memory limits. Run:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
proc options group = macro;
run;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
to verify the following options:
| option | description |
|-------------:|:-----------------------------------------------------------------------------------------------|
|`MEXECSIZE=` | specifies the maximum macro size that can be executed in memory. |
|`MSYMTABMAX=` | specifies the maximum amount of memory available to the macro variable symbol table or tables. |
|`MVARSIZE=` | specifies the maximum size for a macro variable that is stored in memory. |
---
Package contains:
1. macro appendarray
2. macro appendcell
3. macro array
4. macro concatarrays
5. macro deletemacarray
6. macro do_over
7. macro do_over2
8. macro do_over3
9. macro make_do_over
Required SAS Components:
*Base SAS Software*
*SAS package generated by generatePackage, version 20200911*
The SHA256 hash digest for package macroArray:
`085A0F3D544EAF01378BB6C6B4F429123F8BFEEFC76013D1B05DFADFEE3FA661`
---
# Content description ############################################################################################
## >>> `%appendArray()` macro: <<< <a name="appendarray-macro"></a> ############
The `%appendArray()` macro is a macrowrapper
which allows to concatenate two macroarrays
created by `%array()` macro.
By default values of the second macroarray are *not* removed.
Dimensions of the first macroarray are extended.
The `%appendArray()` macro executes like a pure macro code.
### SYNTAX: #####################################################################
The basic syntax is the following, the `<...>` means optional parameters:
~~~~~~~~~~~~~~~~~~~~~~~sas
%appendArray(
first
,second
)
~~~~~~~~~~~~~~~~~~~~~~~
**Arguments description**:
1. `first` - *Required*, a name of a macroarray created by the `%array()` macro.
2. `second` - *Required*, a name of a macroarray created by the `%array()` macro.
### EXAMPLES AND USECASES: ######################################################
**EXAMPLE 1.** Append macroarrays LL and MM.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(ll[2:4] $ 12,
function = quote(put(today() + 10*_I_, yymmdd10.)),
macarray=Y
)
%array(mm[10:13] $ 1200,
function = quote(repeat("A",12*_I_)),
macarray=Y
)
%put *%ll(2)*%ll(3)*%ll(4)*;
%appendArray(ll, mm);
%put *%ll(2)*%ll(3)*%ll(4)*%ll(5)*%ll(6)**%ll(7)*%ll(8)*;
%put *%mm(10)**%mm(11)*%mm(12)*%mm(13)*;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 2.** Error handling.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%appendArray(ll, )
%appendArray(, mm)
%appendArray(noExistA, noExistB)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
---
## >>> `%appendCell()` macro: <<< <a name="appendcell-macro"></a> ##############
The `%appendCell()` macro allows to append
a macrovariable to a macroarray created by the `%array()` macro.
Dimensions of the macroarray are extended.
The `%appendCell()` macro executes like a pure macro code.
### SYNTAX: ####################################################################
The basic syntax is the following, the `<...>` means optional parameters:
~~~~~~~~~~~~~~~~~~~~~~~sas
%appendCell(
first
,second
,hilo
)
~~~~~~~~~~~~~~~~~~~~~~~
**Arguments description**:
1. `first` - *Required*, a name of a macroarray created by the `%array()` macro.
2. `second` - *Required*, a name of a macrovariable to be append to the macroarray.
3. `hilo` - *Required*, if `H` macrovariable is appended at the end
if `L` macrovariable is appended at the beginning
);
### EXAMPLES AND USECASES: ####################################################
**EXAMPLE 1.** Create two macro wrappers.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%* Macro wrapper to append a macrovariable to the end of a macroarray;
%macro appendHC(array,cell);
%appendCell(&array.,&cell.,H)
%mend appendHC;
%* macro wrapper to append a macrovariable to the beginning of a macroarray;
%macro appendLC(array,cell);
%appendCell(&array.,&cell.,L)
%mend appendLC;
%* create macroarrays X and variables W,Y,Z;
%array(X[2:4] $ ("AAA", "BBB", "CCC"), macarray=Y)
%let W=1;
%let Y=2;
%let Z=3;
%put *%do_over(X)*&=W*&=Y*&=Z*;
%put BEFORE *%do_over(X)**&=xLBOUND*&=xHBOUND*&=xN*;
%appendCell(X,Y,H)
%put AFTER1 *%do_over(X)**&=xLBOUND*&=xHBOUND*&=xN*;
%appendLC(X,W)
%put AFTER2 *%do_over(X)**&=xLBOUND*&=xHBOUND*&=xN*;
%appendHC(X,Z)
%put AFTER3 *%do_over(X)**&=xLBOUND*&=xHBOUND*&=xN*;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 2.** Error handling
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%appendCell(X,Y,blahblah)
%appendCell(X,,H)
%appendCell(,Y,H)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 3.** Adding variable below lower bound.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(zero[0:2] $ ("AAA", "BBB", "CCC"), macarray=Y)
%let belowzero=zzz;
%put BEFORE *%do_over(zero)**&=zeroLBOUND*&=zeroHBOUND*&=zeroN*;
%appendCell(zero,belowzero,L)
%put AFTER *%do_over(zero)**&=zeroLBOUND*&=zeroHBOUND*&=zeroN*;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
---
## >>> `%array()` macro: <<< <a name="array-macro"></a> #######################
The code of a macro was inspired by
*Ted Clay's* and *David Katz's* macro `%array()`.
The `%array()` macro version provided in the package
is designed to facilitate
the idea of macro array concept, i.e. *a list of
macrovariables with common prefix and numerical suffixes*.
Usually such construction is then resolved by
double ampersand syntax, e.g. `&&perfix&i` or similar one.
What is new/extension to the `%array()` macro concept are:
0. The syntax is closer to the data step one.
1. It is a pure macro code (it can be executed in any place
of 4GL code), this includes generating macro arrays out
of datasets.
2. When a macroarrray is created it allows also to generate
a new macro (named the same as the array name) and replace
the double ampersand syntax with more array looking one,
i.e. for array ABC user can have `%ABC(1)`, `%ABC(2)`, or `%ABC(&i)`
constructions.
3. The array macro allows to use data step functions to generate
array's entries.
The `%array()` macro executes like a pure macro code.
### SYNTAX: ###################################################################
The basic syntax is the following, the `<...>` means optional parameters:
~~~~~~~~~~~~~~~~~~~~~~~sas
%array(
array
<,function=>
<,before=>
<,after=>
<,vnames=N>
<,macarray=N>
<,ds=>
<,vars=>
)
~~~~~~~~~~~~~~~~~~~~~~~
**Arguments description**:
1. `array` - *Required*, an array name and a declaration/definition of an array, <br>
e.g. `myArr[*] x1-x3 (4:6)` <br>
or `myBrr[*] $ y1-y3 ("a" "b" "c")` <br>
or `myCrr[3] $ ("d d d" "e,e,e" "f;f;f")` <br>
or `myDrr p q r s`. <br>
Macrovariables created by the macro are *global*.
If an array name is `_` (single underscore) then attached variables
list names are used, a call of the form:
`%array(_[*] p1 q2 r3 s4 (-1 -2 -3 -4))`
will create macrovariables: `p1`, `q2`, `r3`, and `s4` with respective
values: `-1`, `-2`, `-3`, and `-4`. <br>
Three additional *global* macrovariables:
`<arrayName>LBOUND`, `<arrayName>HBOUND`, and `<arrayName>N`
are generated with the macroarray. See examples for more use-cases.
* `function=` - *Optional*, a function or an expression to be applied to all array cells,
`_I_` is as array iterator, e.g. `_I_ + rand("uniform")`.
* `before=` - *Optional*, a function or an expression to be added before looping through
array, e.g. `call streaminit(123)`.
* `after=` - *Optional*, a function or an expression to be added after looping through
array, e.g. `call sortn(ABC)`.
* `vnames=N` - *Optional*, default value `N`, if set to `Y`/`YES` then macroarray is built based
on variables names instead values, e.g.
`%array(myArr[*] x1-x3 (4:6), vnames=Y)`
will use `x1`, `x2`, and `x3` as values instead `4`, `5`, and `6`.
* `macarray=N` - *Optional*, default value `N`, if set to `Y`/`YES` then macro named with array
name is compiled to create convenient envelope for multiple ampersands, e.g.
`%array(myArr[*] x1-x3 (4:6), macarray=Y)`
will create `%myArr(J)` macro which will allow to extract "data"
from macroarray like:
`%let x = %myArr(1);`
or when used with second parameter equal `I` (insert) allow to overwrite macroarrays
value:
`%let %myArr(17,i) = 42;`
* `ds=` - *Optional*, use a dataset as a basis for a macroarray data,
if used by default overwrites use of the `array` parameter, honors `macarray=`
argument, dataset options are allowed, e.g. `sashelp.class(obs=5)`
* `vars=` - *Optional*, a list of variables used to create macroarrays from a dataset,
the list format can be as follows (`<...>` means optional):
`variable1<delimiter><arrayname1> <... variableN<delimiter><arraynameN>>`
delimiters are hash(`#`) and pipe(`|`), currently only space
is supported as separator, the meaning of `#` and `|` delimiters
will be explained in the following example:
if the `vars = height#h weight weight|w age|` value is provided
then the following macroarrays will be created: <br>
1) macroarray "H" with ALL(`#`) values of variable "height" <br>
2) macroarray "WEIGHT" with ALL(no separator is equivalent to #)
values of variable "weight" <br>
3) macroarray "W" with UNIQUE(|) values of variable "weight" and <br>
4) macroarray "AGE" with UNIQUE(|) values of variable "age",
### EXAMPLES AND USECASES: ####################################################
**EXAMPLE 1.** Basic use-case.
Creating macroarray like in the array statement;
values are used by default;
different types of brackets are allowed;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(a[*] x1-x5 (1:5))
%array(b{5} (5*17))
%* Mind the $ since it is a character array!;
%array(c(3) $ 10 ("a A" "b,B" "c;C"))
%array(d x1-x5 (5 4 3 2 1))
%put _user_;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 2.** Index ranges.
If range starts < 0 then it is shifted to 0.
In case when range is from `1` to `M`
then macrovariable `<arrayname>N` is set to `M`
In case when range is different
the `<arrayname>N` returns number of
elements in the array `(Hbound - Lbound + 1)`.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(d[-2:2] $ ("a" "b" "c" "d" "e"))
%put &=dLBOUND. &=dHBOUND. &=dN.;
%put &=d0. &=d1. &=d2. &=d3. &=d4.;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 3.** Functions.
It is possible to assign value of a function
or an expression to a cell of the array,
e.g. `array[_I_] = function(...)`.
You can use an iterator in a function.
As in case of usual arrays it is `_I_`.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(e[-3:3] $, function = "A" )
%put &=eLBOUND. &=eHBOUND. &=eN.;
%put &=e0. &=e1. &=e2. &=e3. &=e4. &=e5. &=e6.;
%array(f[-3:3], function = (2**_I_) )
%put &=fLBOUND. &=fHBOUND. &=fN.;
%put &=f0. &=f1. &=f2. &=f3. &=f4. &=f5. &=f6.;
%array(g[0:2], function = ranuni(123) )
%put &=gLBOUND. &=gHBOUND. &=gN.;
%put &=g0. &=g1. &=g2.;
%* Or something more complex;
%array(gg[0:11] $ 11, function = put(intnx("MONTH", '1jun2018'd, _I_, "E"), yymmn.))
%put &=ggLBOUND. &=ggHBOUND. &=ggN.;
%put &=gg0 &=gg1 &=gg2 ... &=gg11;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 4.** Functions cont.
If there is need for set-up something *before* or *after*:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(h[10:12]
,function = rand('Uniform')
,before = call streaminit(123)
,after = call sortn(of h[*])
)
%put &=h10. &=h11. &=h12.;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 5.** Fibonacci series.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(i[1:10] (10*0)
,function = ifn(_I_ < 2, 1, sum(i[max(_I_-2,1)], i[max(_I_-1,2)]) ) )
%put &=i1 &=i2 &=i3 &=i4 &=i5 &=i6 &=i7 &=i8 &=i9 &=i10;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 6a.** "Uppercas Letters"
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(UL[26] $, function = byte(rank("A")+_I_-1) )
%put &=UL1 &=UL2 ... &=UL25 &=UL26;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 6b.** "Lowercase Letters"
Extended by `macarray=Y` option and
the input mode support (with `I`).
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(ll[26] $, function = byte(rank("a")+_I_-1), macarray=Y)
%put *%ll(&llLBOUND.)*%ll(3)*%ll(4)*%ll(5)*...*%ll(25)*%ll(&llHBOUND.)*;
%* The range handling, warning;
%put *%ll(265)*;
%* The input mode;
%put *before:*%ll(2)*;
%let %ll(2,I) = bbbbb;
%put *after: *%ll(2)*;
%* The range handling, error;
%let %ll(265,I) = bbb;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 7.** The use of `vnames=Y`
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(R R1978-R1982)
%put &=R1 &=R2 &=R3 &=R4 &=R5;
%array(R R1978-R1982 (78:82))
%put &=R1 &=R2 &=R3 &=R4 &=R5;
%array(R R1978-R1982 (78:82), vnames=Y)
%put &=R1 &=R2 &=R3 &=R4 &=R5;
%array(R R1978-R1982, vnames=Y)
%put &=R1 &=R2 &=R3 &=R4 &=R5;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 8.** A "no name" array i.e. the `_[*]` array
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(_[*] x1-x5 (1:5))
%put _user_;
%array(_[*] p q r s (4*42))
%put _user_;
%* If no variables names than use _1 _2 ... _N;
%array(_[4] (-1 -2 -3 -4))
%put &=_1 &=_2 &=_3 &=_4;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 9.** Pure macro code can be used in a data step.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
data test1;
set sashelp.class;
%array(ds[*] d1-d4 (4*17))
a1 = &ds1.;
a2 = &ds2.;
a3 = &ds3.;
a4 = &ds4.;
run;
data test2;
set sashelp.class;
%array(_[*] j k l m (4*17))
a1 = &j.;
a2 = &k.;
a3 = &l.;
a4 = &m.;
run;
data test3;
set sashelp.class;
%array(alpha[*] j k l m (101 102 103 104), macarray=Y)
a1 = %alpha(1);
a2 = %alpha(2);
a3 = %alpha(3);
a4 = %alpha(4);
a5 = %alpha(555);
run;
data test4;
set sashelp.class;
%array(beta[*] j k l m (101 102 103 104), vnames=Y, macarray=Y)
a1 = "%beta(1)";
a2 = "%beta(2)";
a3 = "%beta(3)";
a4 = "%beta(4)";
a5 = "%beta(555)";
run;
data test5;
set sashelp.class;
%array(gamma[4] $ 12 ("101" "102" "103" "104"), macarray=Y)
a1 = "%gamma(1)";
a2 = "%gamma(2)";
a3 = "%gamma(3)";
a4 = "%gamma(4)";
a5 = "%gamma(555)";
run;
data test6;
set sashelp.class;
%array(ds = sashelp.cars, vars = Cylinders|, macarray=Y)
a0 = %Cylinders(0);
a1 = %Cylinders(1);
a2 = %Cylinders(2);
a3 = %Cylinders(3);
a4 = %Cylinders(4);
a5 = %Cylinders(555);
run;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 10.** Creating an array from a dataset, basic case.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(ds = sashelp.class, vars = height weight age)
%put _user_;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 11. Creating an array from a dataset, advanced.
If: `vars = height#h weight weight|w age|`
then create:
1. macroarray "h" with ALL(#) values of variable "height"
2. macroarray "weight" with ALL(no separator is equivalent to #) values of variable "weight"
3. macroarray "w" with UNIQUE(|) values of variable "weight"
4. macroarray "age" with UNIQUE(|) values of variable "age"
Currently the only separator in VARS is a space.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(ds = sashelp.class, vars = height#h weight weight|w age|)
%put _user_;
%array(ds = sashelp.class, vars = height#hght weight weight|wght age|, macarray=Y)
%put *%hght(&hghtLBOUND.)**%weight(2)**%wght(&wghtHBOUND.)**%age(3)*;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 12.** Creating an array from a dataset with dataset options
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(ds = sashelp.cars(obs=100 where=(Cylinders=6)), vars = Make| Type| Model, macarray=Y)
%put *%make(&makeLBOUND.)*%Model(2)*%Model(3)*%Model(4)*%type(&typeHBOUND.)*;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
---
## >>> `%concatArrays()` macro: <<< <a name="concatarrays-macro"></a> ###########
The `%concatArrays()` macro allows to concatenate
two macroarrays created by the `%array()` macro.
By default values of the second macroarray are removed.
Dimensions of the first macroarray are extended.
The `%concatArrays()` macro executes like a pure macro code.
### SYNTAX: #####################################################################
The basic syntax is the following, the `<...>` means optional parameters:
~~~~~~~~~~~~~~~~~~~~~~~sas
%concatArrays(
first
,second
<,removeSecond=Y>
)
~~~~~~~~~~~~~~~~~~~~~~~
**Arguments description**:
1. `first` - *Required*, a name of a macroarray created by the `%array()` macro.
2. `second` - *Required*, a name of a macroarray created by the `%array()` macro.
* `removeSecond=Y` - *Optional*, default value `Y`, if set to `Y` then
the second array is removed.
### EXAMPLES AND USECASES: ####################################################
**EXAMPLE 1.** Concatenate macroarrays LL and MM.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(ll[2:4] $ 12,
function = quote(put(today() + 10*_I_, yymmdd10.)),
macarray=Y
)
%array(mm[10:13] $ 12000,
function = quote(repeat("A",123*_I_)),
macarray=Y
)
%put *%ll(2)*%ll(3)*%ll(4)*;
%concatArrays(ll, mm);
%put *%ll(2)*%ll(3)*%ll(4)*%ll(5)*%ll(6)**%ll(7)*%ll(8)*;
%put *%mm(10)**%mm(11)*%mm(12)*%mm(13)*;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 2.** Error handling.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%concatArrays(ll, )
%concatArrays(, mm)
%concatArrays(noExistA, noExistB)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
---
## >>> `%deleteMacArray()` macro: <<< <a name="deletemacarray-macro"></a> #######
The `%deleteMacArray()` macro allows to delete
macroarrays created by the `%array()` macro.
The `%deleteMacArray()` macro executes like a pure macro code.
### SYNTAX: #####################################################################
The basic syntax is the following, the `<...>` means optional parameters:
~~~~~~~~~~~~~~~~~~~~~~~sas
%deleteMacArray(
arrs
<,macarray=N>
)
~~~~~~~~~~~~~~~~~~~~~~~
**Arguments description**:
1. `arrs` - *Required*, a space separated list of manes
of macroarray created by the `%array()` macro.
* `macarray=N` - *Optional*, indicator should a macro
associated with macroarray to be deleted?
If `Y` or `YES` then the associated macro is deleted.
## >>> `%do_over()` macro: <<< <a name="do-over-macro"></a>######################
The code of the macro was inspired by
*Ted Clay's* and *David Katz's* macro `%do_over()`.
The `%DO_OVER()` macro allows to iterate over macroarray created with
the `macarray=Y` parameter of the `%ARRAY()` macro.
The `%do_over()` macro executes like a pure macro code.
### SYNTAX: #####################################################################
The basic syntax is the following, the `<...>` means optional parameters:
~~~~~~~~~~~~~~~~~~~~~~~sas
%do_over(
array
<,phrase=%nrstr(%&array(&_I_.))>
<,between=%str( )>
<,which = >
)
~~~~~~~~~~~~~~~~~~~~~~~
**Arguments description**:
1. `array` - *Required*, indicates a macroarray which metadata (Lbound, Hbouns)
are to be used to loop in the `%do_over()`
* `phrase=` - *Optional*, Default value `%nrstr(%&array(&_I_.))`,
a statement to be called in each iteration
of the internal do_over's loop. Loop iterator is `_I_`,
if you want to use `_I_` or array name
[e.g. `%myArr(&_I_.)`] *enclose it* in the `%NRSTR()`
macro quoting function.
* `between=` - *Optional*, default value `%str( )` (space),
a statement to be called in between each
iteration of the internal do_over loop.
If macroquoted (e.g. `%str( + )`) then the `%unquote()`
function is automatically applied.
* `which=` - *Optional*, a _SPACE_ separated list of indexes which
should be used to iterate over selected macroarray.
Possible special characters are `H` and `L` which means
*high* and *low* bound of an array, list could be set with
colons(`:`) in form of `start:end:by` (*no spaces between!*),
if `by` is omitted the default is `1`. If possible use
`1:5` rather `1 2 3 4 5` since the firs works faster.
### EXAMPLES AND USECASES: ####################################################
**EXAMPLE 1.** Simple looping.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(beta[*] j k l m (101 102 103 104), vnames=Y, macarray=Y)
%put #%do_over(beta)#;
%put #%do_over(beta, phrase=%nrstr("%beta(&_I_.)"), between=%str(,))#;
data test1;
%array(beta[*] j k l m (101 102 103 104), vnames=Y, macarray=Y)
%do_over(beta, phrase=%nrstr(a&_I_. = "%beta(&_I_.)";))
put _all_;
run;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 2.** Multiple arrays looping.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(alpha[*] j k l m n, vnames=Y, macarray=Y)
%array( beta[5] $ , function = "a", macarray=Y)
%array(gamma[4] (101 102 103 104), macarray=Y)
data test2;
call streaminit(123);
%do_over(beta
, phrase = %nrstr(%beta(&_I_.) = %gamma(&_I_.) * rand('Uniform'); output;)
, between = put _all_;
);
put _all_;
run;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 3.** Multiple arrays looping, cont.
Create multiple datasets.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%do_over(beta
, phrase = %nrstr(
data %alpha(&_I_.)2;
call streaminit(123);
%beta(&_I_.)x = %gamma(&_I_.) * rand('Uniform');
output;
run;
)
)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 4.** Multiple arrays looping, cont.
Create multiple datasets using a macro.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%macro doit(ds, var=a, val=1);
data &ds.;
call streaminit(123);
&var. = &val. * rand('Uniform');
output;
run;
%mend doit;
%do_over(beta
, phrase = %nrstr(
%DOIT(%alpha(&_I_.)1, var = %beta(&_I_.), val = %gamma(&_I_.))
)
)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 5.** `%do_over()` inside `%array()`
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(test[*] x1-x12 (1:12), macarray=Y)
%put **%test(1)**%test(12)**;
%put #%do_over(test)#;
%array(abc[*] x1-x12 (%do_over(test,phrase=%nrstr(%eval(100-%test(&_I_.))))), macarray=Y)
%put **%abc(1)**%abc(12)**;
%put #%do_over(abc)#;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 6.** Looping over array with *macroquoted* separator.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(alpha[11] (5:15), macarray=Y)
%let x = %do_over(alpha
, phrase = %NRSTR(%alpha(&_I_.))
, between= %str( + )
);
%put &=x.;
%put %sysevalf(&x.);
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 7.** Working with the `WHICH=` optional parameter
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(test[*] x01-x12, vnames= Y, macarray=Y)
%put #%do_over(test)#;
%put #%do_over(test, which= 1 3 5)#;
%put #%do_over(test, which= 1:5)#;
%put #%do_over(test, which= 1:5:2 7 8)#;
%put #%do_over(test, which= L:H l:h)#;
%put #%do_over(test, which= L:3 10:h)#;
%put #%do_over(test, which= L:H h:l:-1 13 14)#;
%put #%do_over(test, which= %eval(1+1):%eval(5+1))#;
%put #%do_over(test, which= L:H h:l:-1 13 14, between=%str(,))#;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
---
## >>> `%do_over2()` macro: <<< <a name="do-over2-macro"></a>####################
The code of the macro was inspired by
*Ted Clay's* and *David Katz's* macro `%do_over()`.
The `%DO_OVER2()` macro allows to iterate over *two* macroarray created with
the `macarray=Y` parameter of the `%ARRAY()` macro.
The `%do_over2()` macro executes like a pure macro code.
### SYNTAX: #####################################################################
The basic syntax is the following, the `<...>` means optional parameters:
~~~~~~~~~~~~~~~~~~~~~~~sas
%do_over2(
arrayI
,arrayJ
<,phrase=%nrstr(%&arrayI(&_I_.) %&arrayJ(&_J_.))>
<,between=%str( )>
)
~~~~~~~~~~~~~~~~~~~~~~~
**Arguments description**:
1. `arrayI` - Required, indicates the first macroarray which metadata (Lbound, Hbouns)
are to be used in the outer loop in the `%do_over2()`
2. `arrayJ` - Required, indicates the second macroarray which metadata (Lbound, Hbouns)
are to be used in the inner loop in the `%do_over2()`
* `phrase=` - *Optional*, default value `%nrstr(%&arrayI(&_I_.) %&arrayJ(&_J_.))`,
a statement to be called in each iteration
of the *inner* loop. The outer loop iterator is `_I_`,
the inner loop iterator is `_J_`,
if you want to use `_I_`, `_J_`, or arrays names
[e.g. `%myArr(&_I_.)`] *enclose them* in the `%NRSTR()`
macro quoting function.
* `between=` - *Optional*, default value `%str( )` (space),
a statement to be called in between each
iteration of the internal do_over2 loop.
If macroquoted (e.g. `%str( + )`) then the `%unquote()`
function is automatically applied.
### EXAMPLES AND USECASES: ####################################################
**EXAMPLE 1.** Looping over two arrays.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(alpha[*] j k l m n, vnames=Y, macarray=Y)
%array( beta[4] (101 102 103 104), macarray=Y)
%put *%do_over2(alpha, beta
, phrase = %NRSTR((%alpha(&_I_.), %beta(&_J_)))
)*;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 2.** Looping over two arrays with a separator.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(alpha[11] (5:15), macarray=Y)
%array( beta[ 4] (101 102 103 104), macarray=Y)
%let x = %do_over2(alpha, beta
, phrase = %NRSTR((%alpha(&_I_.) * %beta(&_J_)))
, between= +
);
%put &=x.;
%put %sysevalf(&x.);
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 3.** Looping over two arrays with *macroquoted* separator.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(alpha[11] (5:15), macarray=Y)
%array( beta[ 4] (101 102 103 104), macarray=Y)
%let x = %do_over2(alpha, beta
, phrase = %NRSTR((%alpha(&_I_.) * %beta(&_J_)))
, between= %str( + )
);
%put &=x.;
%put %sysevalf(&x.);
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
---
## >>> `%do_over3()` macro: <<< <a name="do-over3-macro"></a>####################
The code of the macro was inspired by
*Ted Clay's* and *David Katz's* macro `%do_over()`.
The `%DO_OVER3()` macro allows to iterate over *three* macroarray created with
the `macarray=Y` parameter of the `%ARRAY()` macro.
The `%do_over3()` macro executes like a pure macro code.
### SYNTAX: #####################################################################
The basic syntax is the following, the `<...>` means optional parameters:
~~~~~~~~~~~~~~~~~~~~~~~sas
%do_over2(
arrayI
,arrayJ
,arrayK
<,phrase=%nrstr(%&arrayI(&_I_.) %&arrayJ(&_J_.) %&arrayK(&_K_.))>
<,between=%str( )>
)
~~~~~~~~~~~~~~~~~~~~~~~
**Arguments description**:
1. `arrayI` - *Required*, indicates the first macroarray which metadata (Lbound, Hbouns)
are to be used in the outer loop in the `%do_over3()`
2. `arrayJ` - *Required*, indicates the second macroarray which metadata (Lbound, Hbouns)
are to be used in the inner loop in the `%do_over3()`
3. `arrayK` - *Required*, indicates the third macroarray which metadata (Lbound, Hbouns)
are to be used in the inner loop in the `%do_over3()`
* `phrase=` - *Optional*, default value `%nrstr(%&arrayI(&_I_.) %&arrayJ(&_J_.) %&arrayK(&_K_.))`,
a statement to be called in each iteration
of the *inner* loop. The *outer* loop iterator is `_I_`,
the *middle* loop iterator is `_J_`, the *inner* loop iterator is `_K_`,
if you want to use `_I_`, `_J_`, `_K_`, or arrays names
[e.g. `%myArr(&_I_.)`] *enclose them* in the `%NRSTR()`
macro quoting function.
* `between=` - *Optional*, default value `%str( )` (space),
a statement to be called in between each
iteration of the internal do_over2 loop.
If macroquoted (e.g. `%str( + )`) then the `%unquote()`
function is automatically applied.
### EXAMPLES AND USECASES: ####################################################
**EXAMPLE 1.** Looping over 3 macroarrays.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(a1_[2] (0 1), macarray=Y)
%array(a2_[2] (2 3), macarray=Y)
%array(a3_[2] (4 5), macarray=Y)
%do_over3(a1_, a2_, a3_
, phrase = %NRSTR(%put (%a1_(&_I_.), %a2_(&_J_), %a3_(&_K_));)
)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 2.** Looping 3 times over a macroarray.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(a[0:2] (0 1 2), macarray=Y)
%do_over3(a, a, a
, phrase = %NRSTR(%put (%a(&_I_.), %a(&_J_), %a(&_K_));)
)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
---
## >>> `%make_do_over()` macro: <<< <a name="make-do-over-macro"></a> ###########
The code of the macro was inspired by
*Ted Clay's* and *David Katz's* macro `%do_over()`.
The `%make_do_over()` macro allows to generate
the `%DO_OVER<n>()` macros. It works *only* for *n>3*!
The `%make_do_over()` macro does *not* executes like a pure macro code.
### SYNTAX: #####################################################################
The basic syntax is the following, the `<...>` means optional parameters:
~~~~~~~~~~~~~~~~~~~~~~~sas
%make_do_over(
size
)
~~~~~~~~~~~~~~~~~~~~~~~
**Arguments description**:
1. `size` - *Required*, indicates the number of dimensions
(i.e. inner loops) of the `%DO_OVER<n>()` macro.
### EXAMPLES AND USECASES: ####################################################
**EXAMPLE 1.** Code of created "4-loop" `%DO_OVER4()` macro
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%macro do_over4(
arrayI1,
arrayI2,
arrayI3,
arrayI4,
phrase=%nrstr(
%&arrayI1(&_I1_.)
%&arrayI2(&_I2_.)
%&arrayI3(&_I3_.)
%&arrayI3(&_I4_.)
),
between=%str( )
);
%local _I1_ _I2_ _I3_ _I4_;
%do _I1_ = &&&arrayI1.LBOUND %to &&&arrayI1.HBOUND;
%do _I2_ = &&&arrayI2.LBOUND %to &&&arrayI2.HBOUND;
%do _I3_ = &&&arrayI3.LBOUND %to &&&arrayI3.HBOUND;
%do _I4_ = &&&arrayI4.LBOUND %to &&&arrayI4.HBOUND;
%if not (
&_I1_. = &&&arrayI1.LBOUND
AND &_I2_. = &&&arrayI2.LBOUND
AND &_I3_. = &&&arrayI3.LBOUND
AND &_I4_. = &&&arrayI4.LBOUND
)
%then %do;%unquote(&between.)%end;%unquote(%unquote(&phrase.))
%end;
%end;
%end;
%end;
%mend do_over4;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 3.** Create a "4-loop" `%DO_OVER4()` macro
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%make_do_over(4);
%array(a1_[2] (0 1), macarray=Y)
%do_over4(a1_, a1_, a1_, a1_
, phrase = %NRSTR(%put (%a1_(&_I1_.), %a1_(&_I2_), %a1_(&_I3_), %a1_(&_I4_));)
)
%put *%do_over4(a1_, a1_, a1_, a1_
, between = *
)*;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 3.** Create a "5-loop" `%DO_OVER5()` macro
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%make_do_over(5);
%array(a1_[2] (0 1), macarray=Y)
%do_over5(a1_, a1_, a1_, a1_, a1_
, phrase = %NRSTR(%put (%a1_(&_I1_.), %a1_(&_I2_), %a1_(&_I3_), %a1_(&_I4_), %a1_(&_I5_));)
)
%put *%do_over5(a1_, a1_, a1_, a1_, a1_
, between = *
)*
;
options nomprint;
data test2;
%do_over5(a1_, a1_, a1_, a1_, a1_
, phrase = %NRSTR(x1 = %a1_(&_I1_.); x2 = %a1_(&_I2_); x3 = %a1_(&_I3_); x4 = %a1_(&_I4_); x5 = %a1_(&_I5_);)
, between = output;
)
output;
run;
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**EXAMPLE 4.** Create all from 6 to 10 "do_overs"
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sas
%array(loop[6:10] (6:10), macarray=Y)
%do_over(loop
, phrase = %nrstr(
%make_do_over(%loop(&_I_.))
)
);
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
---
## License ####################################################################
Copyright (c) Bartosz Jablonski, since January 2019
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
---