Perl Weekly Challenge 170: primordial matrix!
It is sad that, after more than two years of me doing Raku, I still don’t have any production code project to work on. Therefore, in order to keep my coding and Raku-ing (is that a term?) knowdledge, I try to solve every Perl Weekly Challenge tasks.In the following, the assigned tasks for Challenge 170.
and for the sake of some Perl 5, let’s do some stuff also in PostgreSQL Pl/Perl:
Last, the solutions in PostgreSQL PL/PgSQL:
PWC 170 - Task 1
Compute the first ten primordial numbers, those that are made up of multiplication of prime numbers. This is a piece of cake in Raku:sub MAIN( Int $limit = 10 ) {
my @primes;
my @primordial = lazy gather {
for ( 1 .. Inf ) {
next if ! .is-prime;
@primes.push: $_;
take [*] @primes;
}
};
@primordial[ 0 .. $limit ].join( "\n" ).say;
}
There is a
lazy-gather that populates the @primes array and returns the reduction of the multiplication of all its values.
In order to speed up the computation, once could also cache the result of the last computation and multiply it by the next prime in order to not consume both space and time.
PWC 170 - Task 2
A strange matrix product, named Kronecker product. Assuming both the matrixes have the same shape, this is also another piece of cake in Raku:sub MAIN() {
my @a = [1, 2], [3, 4];
my @b = [5, 6], [7, 8];
my @result;
for 0 ..^ @a.elems -> $row_a {
for 0 ..^ @b.elems -> $row_b {
@result.push: [ @a[ $row_a ].List X* @b[ $row_b ].List ];
}
}
@result.join( "\n" ).say;
}
I extract every line out of a matrix and multiply it with the corresponding line on the other matrix, bia the
X* cross-multiply operator. The result is placed into the @result matrix.
PWC 170 - Task 1 in PostgreSQL PL/Perl
Very likely the Raku solution:create table if not exists a( a int, b int );
truncate table a;
insert into a values (1,2), (3,4);
create table if not exists b( a int, b int );
truncate table b;
insert into b values (5,6), (7,8);
CREATE SCHEMA IF NOT EXISTS pwc170;
CREATE OR REPLACE FUNCTION
pwc170.task2_plperl( text, text )
RETURNS TABLE( a int, b int, c int, d int )
AS $CODE$
my ( $table_a, $table_b ) = @_;
elog( DEBUG, "Reading tables $table_a and $table_b" );
my ( $rs_a, $rs_b );
$rs_a = spi_exec_query( "SELECT a,b FROM $table_a" );
$rs_b = spi_exec_query( "SELECT a,b FROM $table_b" );
for my $row_a ( 0 .. $rs_a->{ processed } - 1 ) {
elog( DEBUG, "Loop A $row_a out of " . $rs_a->{ processed } );
my ($aa, $ab) = ( $rs_a->{ rows }[ $row_a ]->{ a }, $rs_a->{ rows }[ $row_a ]->{ b } );
for my $row_b ( 0 .. $rs_b->{ processed } - 1 ) {
elog( DEBUG, "Loop B $row_b out of " . $rs_b->{ processed } );
my ($ba, $bb) = ( $rs_b->{ rows }[ $row_b ]->{ a }, $rs_b->{ rows }[ $row_b ]->{ b } );
elog( DEBUG, "Computing $aa $ab X* $ba $bb" );
my $result = {
a => $aa * $ba,
b => $aa * $bb,
c => $ab * $ba,
d => $ab * $bb,
};
return_next( $result );
}
}
return undef;
$CODE$
LANGUAGE plperl;
Tables
a and b are initialized with the correct values.
The function takes as input arguments the names of the table to query, and do a nested loop on every single row extracted from each table. Columns are assigned to four different variables that are then multiplied togheter and stored into an hash that plays the role of the output table. Then the result is appended to the result set.
PWC 170 - Task 2 in PostgreSQL PL/Perl
A little more complicated solution, that is based on the fact that the matrixes are now tables:CREATE OR REPLACE FUNCTION
pwc170.task2_plperl( int )
RETURNS SETOF int
AS $CODE$
use Math::BigInt;
my ($limit) = @_;
my $is_prime = sub {
my ($value) = @_;
for ( 2 .. $value - 1 ) {
return 0 if $value % $_ == 0;
}
return 1;
};
my $compute_factors = sub {
my ($value) = @_;
my @factors;
for ( 2 .. $value - 1 ) {
next if ! $is_prime->( $_ );
while ( $value % $_ == 0 ) {
push @factors, $_;
$value /= $_;
}
}
return @factors;
};
my $min = sub {
my $found = shift @_;
for ( @_ ) {
$found = $_ if $_ < $found;
}
return $found;
};
my $is_achille = sub {
my ($number) = @_;
my $bag = {};
for ( $compute_factors->( $number ) ) {
$bag->{ $_ }++;
}
return $min->( values( %$bag ) ) >= 2 && Math::BigInt::bgcd( values( %$bag ) )->numify == 1;
};
for ( 1 .. 999999 ) {
if ( $is_achille->( $_ ) ) {
$limit--;
return_next( $_ );
}
last if ! $limit;
}
return undef;
$CODE$
LANGUAGE plperlu;
I could have used other modules to get prime numbers and factors, but I preferred to keep as simpler as possible the solution.
Here implementing the
Bag class behavior is easy thanks to autovivification!
PWC 170 - Task 1 in PostgreSQL PL/PgSQL
This approach does the caching I was writing about in the Raku implementation:CREATE OR REPLACE FUNCTION
pwc170.is_prime( v bigint )
RETURNS bool
AS $code$
DECLARE
i bigint;
BEGIN
FOR i IN 2 .. v - 1 LOOP
IF v % i = 0 THEN
RETURN false;
END IF;
END LOOP;
RETURN TRUE;
END
$code$
LANGUAGE plpgsql;
CREATE OR REPLACE FUNCTION
pwc170.task1_plpgsql( l int default 10 )
RETURNS SETOF INT
AS $CODE$
DECLARE
v bigint := 1;
i bigint;
BEGIN
FOR i IN SELECT n FROM generate_series( 1, 100000 ) n LOOP
IF pwc170.is_prime( i ) THEN
v := v * i;
l := l - 1;
RETURN NEXT v;
END IF;
IF l <= 0 THEN
RETURN;
END IF;
END LOOP;
RETURN;
END
$CODE$
LANGUAGE plpgsql;
There is a
is_prime utility function, but in the main loop I do keep multiplying the previous result by the new computed prime, returning it into the result set. Therefore, there is no array of primes to iterate onto.
PWC 170 - Task 2 in PostgreSQL PL/PgSQL
SQL has already such computation: cross join!However, let’s add something more useful on top of it: the capability to specify from which tables extract the resulting matrix. In this example I’m going to use the same tables
a and b of the previous PL/Perl implementation:
CREATE OR REPLACE FUNCTION
pwc170.task2_plpgsql( ta text, tb text )
RETURNS TABLE ( a int, b int, c int, d int )
AS $CODE$
DECLARE
BEGIN
RETURN QUERY
EXECUTE format( 'SELECT a.a * b.a, a.a * b.b, a.b * b.a, a.b * b.b FROM %I a, %I b',
ta, tb );
END
$CODE$
LANGUAGE plpgsql;
The function takes as input arguments the names of the tables to use as matrixes. Then it formats a query that does the cross join with the correct names and returns the query as a whole.