On Sat, Jun 26, 2010 at 3:47 PM, P Kishor <[email protected]> wrote: > On Sat, Jun 26, 2010 at 3:39 PM, Chris Marshall <[email protected]> wrote: >> On 6/26/2010 1:55 PM, P Kishor wrote: >>> >>> a few days back I had asked for help creating an [x,y,z] dims piddle >>> where every element was a random integer between 0 and 255. A couple >>> of suggestions set me off to learn dummy(), and I also learned the >>> correct use of random() (thanks, Daniel and Chris). Today, I figured >>> out a way to make any dimension piddle filled with random integers >>> between any two integers<max> and<min>. >>> >>> $a = (( (zeroes(<dim> [,<dim> ..] ))->random * (<max> -<min> + 1 >>> )) +<min> )->floor >>> >>> The thing is, the above works whether I use zeroes or ones, which, I >>> am guessing is because internally PDL::new_from_specification is being >>> used to create a template for the piddle. >> >> It will also work if you leave out the ones or zeros/zeroes and >> put the type and dimension arguments in random: >> >> PDL> $a = (random(5,5,3)*(1000-500+1)+500)->floor > > very nice, thanks! That said, I kinda like method calls on objects. I > like $p->random more than random(), and for that reason, I have warmed > to $p->slice more than the NiceSlice syntax. zeroes gives me a piddle > to work on. > >..
ha! now that I have stared at the above, I realize that I was talking nonsense. Your method is absolutely much better than mine because your method requires less typing, hence, less chance of errors, and produces just as good a result. zeroes() and random() are analogous. -- Puneet Kishor _______________________________________________ Perldl mailing list [email protected] http://mailman.jach.hawaii.edu/mailman/listinfo/perldl
