collatz-conjecture by Average-user · Pull Request #593 · exercism/haskell

Average-user · 2017-09-19T00:52:54Z

I'm implementing the collatz conjecture of the Unimplmented excercises of Haskell (excercism.io). But i wasn't sure of what to put on the config.json, uuid part of the excercise.

petertseng · 2017-09-19T01:14:21Z

config.json, uuid part of the excercise.

You guessed correctly (I think) that you generate any arbitrary UUID and use it.

I notice that https://github.com/exercism/docs/blob/master/you-can-help/implement-an-exercise-from-specification.md#configuring-the-exercise did not tell you you are supposed to generate a UUID. That is a deficiency in the documentation, which should be reported.

Please notify if anything else was unclear!

petertseng · 2017-09-19T01:14:52Z

+
+## Source
+
+All of Computer Science [http://www.wolframalpha.com/input/?i=binary&a=*C.binary-_*MathWorld-](http://www.wolframalpha.com/input/?i=binary&a=*C.binary-_*MathWorld-)


this is not the source listed in https://github.com/exercism/problem-specifications/blob/master/exercises/collatz-conjecture/metadata.yml

ups, sorry. Im changing it

petertseng · 2017-09-19T01:16:21Z

@@ -0,0 +1,21 @@
+
+name: collatz-conjecture
+version: 1.0.0.3


as https://github.com/exercism/problem-specifications/blob/master/exercises/collatz-conjecture/canonical-data.json has version 1.1.1, the version that would comply with our versioning policy in #522 is 1.1.1.1, and ensuring that the tests in the Haskell test file match those in the JSON file

(or, alternatively, if you believe those tests are unsuitable for the Haskell track for any reason, let's discuss why they are unsuitable, and then use version0.1.0.1)

petertseng · 2017-09-19T01:43:08Z

@@ -0,0 +1,20 @@
+name: collatz-conjecture
+version: 1.1.1


needs an additional .1 serial number, resulting in 1.1.1.1

Average-user · 2017-09-19T01:46:34Z

Well. I'm glad to have guessed. But i still without understand Travis, Will ever let me free?

Average-user · 2017-09-19T02:29:11Z

Travis gives me an error of another excercise

Registering accumulate-0.0.0...
accumulate-0.0.0: test (suite: test)
empty accumulation
accumulate squares
accumulate upcases
accumulate reversed strings
accumulate recursively
accumulate non-strict FAILED [1]
Failures:
  test/Tests.hs:35: 
  1) accumulate non-strict
       uncaught exception: ErrorCall (accumulate should be even lazier, don't use reverse!
       CallStack (from HasCallStack):
         error, called at test/Tests.hs:36:45 in main:Main)
Randomized with seed 1712205088
Finished in 0.0005 seconds
6 examples, 1 failure
Completed 2 action(s).
Test suite failure for package accumulate-0.0.0
    test:  exited with: ExitFailure 1
Logs printed to console```

petertseng · 2017-09-19T03:31:24Z

Travis gives me an error of another excercise

testing /home/travis/build/exercism/haskell/exercises/accumulate/examples/fail-reverse - should build, but fail tests

It should fail, and it did, so no problem. Please look for problems related to collatz-conjecture

Average-user · 2017-09-19T03:34:42Z

But why still a red X on my pull request?

And i was thinking that Knuth shuffle algorithm, might be a good excercise

petertseng · 2017-09-19T07:30:55Z

But why still a red X on my pull request?

I don't know, what does the Travis CI output at https://travis-ci.org/exercism/haskell/builds/277117302 say? Alternatively, what do you get if you try to stack test in the collatz-conjecture directory on your computer? Alternatively, what do you get if you try to run https://github.com/exercism/haskell/blob/master/bin/test-all-examples ? Something like bin/test-all-examples exercises/collatz-conjecture ?

And i was thinking that Knuth shuffle algorithm, might be a good excercise

Okay, it can be suggested in https://github.com/exercism/problem-specifications . My advice is that based on exercism/problem-specifications#761 exercises that test a specific algorithm are difficult to write tests for because there is no good way to verify that a student's implementation uses that specific algorithm, rather than any other algorithm that achieves the same effect.

Average-user · 2017-09-19T16:08:28Z

Yeah you are right about the Knuth shuffle algorithm, and it's even worts if you take on count that is a SHUFFLE algorithm, not always returns the same result for the same input (well, strictly yes but ...)

Average-user · 2017-09-19T16:54:52Z

At the end the problem was that i misspelled a file name.

And i followed your advice and wrote an issue on: exercism/problem-specifications#906.
But it's a different exercise, not the Knuth algorithm.

petertseng · 2017-09-19T16:57:52Z

+               , expected    = 0
+               }
+        , Case { description = "collatz 2 require 1 step."
+               , number      = 2


on my advice, you did change the problem version to 1.1.1.1, to match the 1.1.1 of https://github.com/exercism/problem-specifications/blob/master/exercises/collatz-conjecture/canonical-data.json, but you did not then use the test cases in https://github.com/exercism/problem-specifications/blob/master/exercises/collatz-conjecture/canonical-data.json. Can you do one of the two:

Use the test cases that appear in https://github.com/exercism/problem-specifications/blob/master/exercises/collatz-conjecture/canonical-data.json

Confirm that the test cases in https://github.com/exercism/problem-specifications/blob/master/exercises/collatz-conjecture/canonical-data.json are unsuitable for the Haskell track, and thus we cannot use that file, and thus we must change our version to 0.1.0.1

so it's ok if i change it from 1.1.1.1 to 1.1.1 and use this cases ?

1.1.1 is not a valid Haskell track version. The two choices are:

Use 1.1.1.1 and the cases in https://github.com/exercism/problem-specifications/blob/master/exercises/collatz-conjecture/canonical-data.json

Use 0.1.0.1 and whatever cases desired, with an explanation of why https://github.com/exercism/problem-specifications/blob/master/exercises/collatz-conjecture/canonical-data.json cannot be used.

petertseng

I think everything is looking good. Please expect a next review in 12 hours, but until then I have small items you could do

petertseng · 2017-09-19T18:49:26Z

+        , Case { description = "even and odd steps"
+               , property    = "steps"
+               , number      = 12
+               , expected    =Just 9


add a space after the = here?

petertseng · 2017-09-19T18:50:12Z

+      "slug": "collatz-conjecture",
+      "core": false,
+      "unlocked_by": null,
+      "difficulty": 1,


since this exercise now uses Maybe, please add it to the topics list (see "rna-transcription" for example of how it is done)

petertseng · 2017-09-19T18:51:30Z

+                  | otherwise = collatzHelper (t+1) (x*3 + 1)
+
+collatz :: Integer -> Maybe Integer
+collatz x = if  x <= 0 then Nothing else collatzHelper 0 x


there are two spaces after the if, you could remove one of them, right?

petertseng · 2017-09-19T18:52:15Z

+
+cases :: [Case]
+cases = [ Case { description = "zero steps for one"
+               , property    = "steps"


the property is not being used. It should be removed.

i will, but maybe it has to be removed from https://github.com/exercism/problem-specifications/blob/master/exercises/collatz-conjecture/canonical-data.json

It needs to be in problem-specifications since it tells what is being tested.

It must not be here since we already know what we are testing (it's specified in code)

petertseng

I think this is good. Please expect me to merge it in 4 hours, unless you see anything you want to change.

Average-user · 2017-09-20T01:52:28Z

No, do it. I'm ok with that

petertseng · 2017-09-20T04:48:08Z

Thank you for implementing this exercise!

petertseng · 2017-09-20T06:02:36Z

@@ -0,0 +1,52 @@
+{-# OPTIONS_GHC -fno-warn-type-defaults #-}


I don't think this line is necessary. Can you check? If you confirm it is not necessary, please send a PR that deletes it.

I retract. I think #144 tells me it is desirable to have for exercises with numbers.

Average-user added 4 commits September 18, 2017 20:56

New excercise. collatz-conjecture

ca1c340

collatz-conjecture added to config.json

319fdfa

adding collatz-conjecture

b248147

collatz conjecture

d353964

petertseng reviewed Sep 19, 2017

View reviewed changes

Lucas Polymeris added 4 commits September 18, 2017 22:25

Update README.md

8d850ca

Update package.yaml

e99495f

Update package.yaml

80f3376

Update package.yaml

4f97913

petertseng reviewed Sep 19, 2017

View reviewed changes

Update package.yaml

c8cfb11

Update Tests.hs

611ed61

name fixed

492764c

petertseng reviewed Sep 19, 2017

View reviewed changes

Average-user added 3 commits September 19, 2017 15:02

new cases

d8c1170

handling errors with Maybe

2ed3db6

fix on src

9061aeb

petertseng reviewed Sep 19, 2017

View reviewed changes

a few corrections

14dfaeb

petertseng approved these changes Sep 20, 2017

View reviewed changes

collatz: regenerate readme

6c6484a

petertseng merged commit 7dc8cf2 into exercism:master Sep 20, 2017

petertseng reviewed Sep 20, 2017

View reviewed changes


		## Source

		All of Computer Science [http://www.wolframalpha.com/input/?i=binary&a=C.binary-_MathWorld-](http://www.wolframalpha.com/input/?i=binary&a=C.binary-_MathWorld-)

Uh oh!

Conversation

Average-user commented Sep 19, 2017

Uh oh!

petertseng commented Sep 19, 2017

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Average-user commented Sep 19, 2017

Uh oh!

Average-user commented Sep 19, 2017

Uh oh!

petertseng commented Sep 19, 2017

Uh oh!

Average-user commented Sep 19, 2017

Uh oh!

petertseng commented Sep 19, 2017

Uh oh!

Average-user commented Sep 19, 2017

Uh oh!

Average-user commented Sep 19, 2017

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

petertseng left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

petertseng left a comment

Choose a reason for hiding this comment

Uh oh!

Average-user commented Sep 20, 2017

Uh oh!

petertseng commented Sep 20, 2017

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants