On 4/22/2014 3:17 AM, Norbert Melzer wrote:

What Is the the type of your pappus function?

I've not specified a type sig, so it figures out some particular subdomain of numeric types. Whatever that is, it takes two numbers and returns a tuple of three floating-point numbers pappus r n = (x-0.5,y,rn) where blah= n**2*(1-r)**2+r x = (r*(1+r)) / (2*blah) y = (n*r*(1-r)) / blah rn = ((1-r)*r)/(2*blah)

On 2014-04-22 10:49, Norbert Melzer wrote:

This one doesn't work, since pappus doesn't return a list

Ah, right, I forgot to adjust that part. How about chain1 = [ translate x y $ color red $ Circle r | i <- [1..10], let (x,y,r) = pappus 100 i ] ...though I'm not too fond of 'let' declarations in list comprehensions. -- Frerich Raabe - raabe@froglogic.com www.froglogic.com - Multi-Platform GUI Testing

On 4/22/2014 3:47 AM, Frerich Raabe wrote:

chain1 = [ translate x y $ color red $ Circle r | i <- [1..10], (x,y,r) <- pappus 100 i ]

would be one way to do what you want.

Yes, that's along the lines of what I was thinking — the 1..10 needs to be bound by itself to trigger the looping behavior, but didn't know how to then continue to generate (x,y,r). I thought the comma there was followed by a "guard", which constrains which values of i are taken or skipped. Ah, you have a <- there too. Hmm, but normally a second <- is another iteration that's done first/loops faster. So what's the syntax here? I must be getting things mixed up. —John

chain1 = [ translate x y $ color red $ Circle r | (x,y,r) <- [pappus 100 i | i <- [1..10]] ] better chain1 = [transform $ pappus 100 i | i <- [1..10]] where transform (x,y,r) = translate x y $ color red $ Circle r 2014-04-22 16:58 GMT+03:00 Kim-Ee Yeoh :

On Tue, Apr 22, 2014 at 3:00 PM, John M. Dlugosz

...

wrote:

...

chain1 = [ translate x y $ color red $ Circle r | (x,y,r) <- pappus 100 [1..10] ]

What is translate? What is color? What is Circle? What is pappus?

None of this is plain haskell.

John, if you make your readers guess at undefined names, they'll go away and hangout somewhere friendlier!

-- Kim-Ee

_______________________________________________ Beginners mailing list Beginners@haskell.org http://www.haskell.org/mailman/listinfo/beginners

Its considered unfriendly to not give important information. The minimum information would have been the typesignature of every function you use. Am 22.04.2014 20:38 schrieb "John M. Dlugosz" :

On 4/22/2014 8:58 AM, Kim-Ee Yeoh wrote:

...

On Tue, Apr 22, 2014 at 3:00 PM, John M. Dlugosz < ngnr63q02@sneakemail.com mailto:ngnr63q02@sneakemail.com> wrote:

chain1 = [ translate x y $ color red $ Circle r | (x,y,r) <- pappus 100 [1..10] ]

What is translate? What is color? What is Circle? What is pappus?

None of this is plain haskell.

John, if you make your readers guess at undefined names, they'll go away and hangout somewhere friendlier!

-- Kim-Ee

Sorry — I thought showing a form that did work would be enough. The important part is that I have a input = map foo [0..10] and a bar (t1,t2,t3) = baz where baz returns a tuple, and the result of bar is the guts of another map.

I didn't realize that using real words from a library instead of foo and bar was considered unfriendly!

_______________________________________________ Beginners mailing list Beginners@haskell.org http://www.haskell.org/mailman/listinfo/beginners

On Wed, Apr 23, 2014 at 1:37 AM, John M. Dlugosz wrote:

I didn't realize that using real words from a library instead of foo and bar was considered unfriendly!

As everyone else has chimed in, signatures + succinct description is a minimum. Haskell usage is very, very diverse. You can't assume that everyone knows the specifics of a library you're using. E.g. graphics usage is scattered all over the place. Many (most?) haskellers hardly meddle with graphics in the slightest. -- Kim-Ee