So what does your code return for encode("")? According to the error message you posted, it returns an index out of range error. Fix that.

I know I'm 2 years late, but I saw your comment and noticed you still don't have a solution.

I found it helps to understand the algorithm if you do some examples by hand.
If you encode "Mellow Yellow" and "Melllow Yelow" by hand, you should see a difference.

Very nice kata. Been using these things for a while and it was a lot of fun / very insightful to see how they're made :)

My return is two elements. I am still getting this error. Can you point me in the right direction?

Even the test is showing one return element

OP solved it, closing

added some time ago

No more help needed, I found the origin.

I am testing Huffman encoding in Python: Passed tests : 2781 but Failed tests: 265

I don't undestand the following error
Log
frequencies(l)
l
encode([],l)
Test Passed
ERROR restore sorted sequence from frequencies should succeed: [] should equal ['l']

Can you help me Please

This kata's basic tests are incredibly inconsistent

test.assert_equals( encode( fs, []), '')
test.assert_equals( encode( [], "" ), None );
test.assert_equals( encode( [('a', 1)], "" ), None );
test.assert_equals( encode( [ ("a",4), ("b",1), ("c",2) ], "" ), "" )

1. Why is it that we are testing enode(fs, []), when according to the hint the second argument should be a string ???

takes: [ (str, int) ], str; returns: String (with "0" and "1")

2. Why is encode(fs, "") supposed to return None if len(fs)<2 and else "" ????

Other people have mentioned the empty string. I was pulling my hair out trying to optimize my code until I realized the timeout I was getting was because I was't handling it. Perhaps handing this as input could be clearer? Maybe that is just my lack of experience.

there is no such thing as an empty array in C, it's forbidden by the standard

Currently working my way through this (C#) and noticed a few things that irk me:

• Sample tests aren't unit tests; for example Foldr requires both Repeat and Take to be implemented. If I'm working my way from top to bottom, then I will not have implemented Take yet, and will receive a slightly confusing NotImplementedError for the Foldr test.
• Sample tests don't properly test the streams: I successfully passed all sample tests, then got an immediate Stack Overflow when I hit "attempt"; turns out a few of my methods were returning (headS, s) instead of (headS, tailS), i.e. recursively running on the same unmodified stream instead of on the tail of the input stream.
• The interface of LazyStream is not specified and only discoverable by chance by reading through the sample tests. Otherwise, I could never have known that LazyStream stores the Stream in its .Value property, and I wouldn't be able to implement anything.

encode? easy :)
decode? 0_0, how I get banana back?

I was getting "SEGV on unknown address" many many times...
until I figured out that input in decode() can be empty string.
hope my experience will help someone in future.

very interesting kata. I should google more about BWT logic, to implement decode()
however it was cool and exiting. thanks for kata.

Exactly, some expected outcomes would be good!