Shame I can't jump to your solution to vote it up, not only it's really clever it's much faster than top voted and mine.
To be exact most "clever" solution is 9 (!) times slower than yours.
Hate the way Codewars promotes writing functions in slow wank style as "Clever Best Practices" :/

Result without benchmark setup is rather meaningless.

Not sure if using Stream is a good idea. I did some benchmark and exec time up ~70% comparing to Enum.filter. So I would rather vote Clever :)

Agree on 3ky, quite easy. I'll get to choppa, thanx :)

This algorithm actually gets exponentially slower as the number increases, so you can't really compare its speed in a 1:1 manner.

2.4.1 :061 > Benchmark.measure { pascalsTriangle(25) }
=> @real=0.000426109

2.4.1 :066 > Benchmark.measure { pascalsTriangle(125) }
=> @real=0.09114601

2.4.1 :062 > Benchmark.measure { pascalsTriangle(250) }
=> @real=3.03741528

2.4.1 :067 > Benchmark.measure { pascalsTriangle(500) }
=> @real=21.5128502

Took me a while looking through the Array docs to find product, I was trying hacky ways with permutation before that.

This code is better than my own though!

This solution is 30 times slower than mine :) Indeed to understand recursion you have to understand recursion first ;)

After submiting I benchmarked this solution against top 10 solutions and this one is the killer. Although my other solution is nifty :)

Indeed. This code is too complex and slow. It may seems "clever", but its definitely not BP

So slow ... I came up with two solutions for this prob, one-string-but-slow solution and fast one. This solution is 2 times slower than my slow :)

Clever solution. I wish I knew product method before.

Elegant solution! Even kind of has a philosophical ring to it :)

Nice one. I've got mine from same string of Test source code :)

The most interesting solution. Mine is also not using eval, but quite simple and easy to read. @JoshBrodieNZ your solution is universal, man, you could use it for half of katas with minor changes ;)