01:22 <Duns_Scrotus> what happens when you jump back into a parameterize form with a continuation

03:37 <ym555> the values of the whatever are reset back to whatever was in the parameterize form

03:37 <ym555> (i think)

03:43 <ym555> http://pasterack.org/pastes/49418

03:48 <ym555> oh whoops replace "whatever" with "parameters" in my sentence

04:28 <NeoHamled> Is empty considered more idiomatic Racket than '()?

04:31 <NeoHamled> Ruby, always helpful with a variety of options, allows for [], Array.new, and even nil.to_a but I believe [] is considered the idiomatic way to write it

05:03 <lexi-lambda> NeoHamled: I think '() is significantly more common in Racket code than empty.

05:07 <NeoHamled> Okay thanks

09:40 <keep_learning_M> Hi Everyone

09:41 <keep_learning_M> I am trying to write a simple gate evaluator which has some contracts https://gist.github.com/mukeshtiwari/fec53c85dffb7733066e9d21ca9bba82#file-contract-rkt-L2

09:41 <keep_learning_M> But I am getting error https://gist.github.com/mukeshtiwari/fec53c85dffb7733066e9d21ca9bba82#file-contract-rkt-L58

09:42 <keep_learning_M> Any hint/idea would be great.

09:53 <rain1> keep_learning_M: the nand-gate structure can only have #t or #f in it

09:53 <rain1> but you need to put other nand-gates in it too

09:54 <keep_learning_M> rain1, Yes, and I am wondering about syntax. I assume it should (or boolean? nand-expr?) but I am not sure.

10:09 <keep_learning_M> So far I have figured out that it should be (or/c boolean? _______). I still need to find what should be at ______.

10:39 <keep_learning_M> Some more experiments led me to understand that (define-struct/contract nand-gate

10:39 <keep_learning_M> ([xin (or/c boolean? nand-gate?)]

10:39 <keep_learning_M> [yin (or/c boolean? nand-gate?)])

10:39 <keep_learning_M> #:transparent)

10:39 <keep_learning_M> is not possible "cannot reference an identifier before its definition"

10:41 <keep_learning_M> Now my question is how can we achieve this kind of constraint of data structure which takes some base type (int, bool) or it self (recursively defined) ?

10:55 <keep_learning_M> Please leave your message in gist. Thank you very much in advance.

13:22 <m4110c> Hi all, I have a little problem and can't find a solution. My problem is, that I have a list of numbers and I wannt to add them all up and calculate the average of them, recursively. So if I had (list 3 6 9), I would calculate the sum, which is 18 and divide by the length of the list which is 3. But on every recursion step the length of the list decreases and I have no "counter" or so to keep track of the

13:22 <m4110c> initial length of the list. I'd really appreciate any hint in the right direction

13:30 <pierpal> make the recursion step such that you have the lenght of the list examined so far.

13:31 <m4110c> pierpal: let me think about it...

13:31 <m4110c> I'll upload my code so far

13:31 <pierpal> k

13:32 <m4110c> http://pasterack.org/pastes/60520

13:33 <m4110c> http://pasterack.org/pastes/48654

13:33 <m4110c> this is better

13:34 <m4110c> removed some "play around stuff"

13:34 <m4110c> The recursion step would be the else clause right?

13:37 <eush> m4110c: you don't need to pass the original length, at each step you just need to (1) compute the sum and (2) divide it by the number of students (length students)

13:38 <eush> m4110c: in order to compute (1) you add up the age of (first student) to the (* (length (rest students)) (mean-of-ages (rest students)))

13:38 <m4110c> eush: but where in the code do I put the division. Because if I put it in the else clause, it gets evaulated on every recurstion step

13:38 <m4110c> and in each step the length is one less than the time before

13:39 <eush> m4110c: if you want to do the division once, why not split it into a recursive function that calculates the sum, and call it from a non-recursive functions that calculates means?

13:39 <eush> (define (means-of-ages students) (/ (sum-of-ages students) (length students)))

13:40 <m4110c> eush: yes, that would be one solution. using a helper function. but I wanted to try without. It's just to help my understanding and play around.

13:41 <eush> m4110c: well, you can also add the counter parameter to your function, and increment it on each recursive step

13:41 <m4110c> I wonder if there is no "mathematical" way to solve it within one function. Like adding something to the sum on each step and dividing at the end... something dependent on the length of the initial list or so. Don't know if I express myself clearly :-D

13:42 <m4110c> Ok, I'll try more and maybe get back to you guys. Thanks for the input so far

13:42 <eush> m4110c: there certainly is a mathematical way to derive the mean of a list from the mean of the sublist

13:44 <eush> m4110c: if your function returns the mean of a list, then each recusion invocation has to have a division in there, so in the end you will perform the division multiple times - i guess there's no getting around this constraint with one function

13:45 <m4110c> yes. let's say the length is n. On every step it becomes n-1. I need to go on from there somehow

13:46 <m4110c> Multiplying or dividing with regard to n on each step and then at the end doing something (which I don't know yet)

13:54 <m4110c> got it!

13:54 <m4110c> see:

13:54 <m4110c> http://pasterack.org/pastes/75756

13:55 <m4110c> I multiply (the length of the list - 1) to every recursion step and divide by (the length of the list) at the end

13:58 <eush> m4110c: you ended up with my suggestion :)

13:59 <m4110c> eush: yep. just the concrete implementation was missing. I had a feeling, that it works mathematically by dividing multiple times. just did not know where exactly in the code to put it.

13:59 <m4110c> I'm not even sure, that I understand it completely, yet. Need to look at it more.

13:59 <m4110c> Thanks for you help! Really great channel here

14:01 <m4110c> Is there something like a "racket visualiser" that lays out the recursive functions to see the full execution path?

19:01 <bremner> rudybot: later tell m4110 it's not graphical, but you might like trace

19:01 <rudybot> bremner: I asked `MemoServ' to forward the message to m4110.