These are chat archives for ramda/ramda

14th
Sep 2017
Simon Friis Vindum
@paldepind
Sep 14 2017 05:57
@buzzdecafe That's cool. I've never seen Black Box before. But it looks like something that could be a lot of fun for a clever child (or grown up) :smile:
Isaac
@IsaacRaja_twitter
Sep 14 2017 06:04
Can someone tell me how ap works?
R.ap([R.multiply(2), R.add(3)], [1,2,3]); //=> [2, 4, 6, 4, 5, 6]
R.ap([R.concat('tasty '), R.toUpper], ['pizza', 'salad']); //=> ["tasty pizza", "tasty salad", "PIZZA", "SALAD"]
This is from ramda documentation
I understand it applies the list of functions to a data set
but how did the results of each function application is concatenated
and how's it different from converge
Adam Szaraniec
@mimol91
Sep 14 2017 06:34

@IsaacRaja_twitter
I've got this same thoughts. (I am not ramda expert so I may be wrong)
converge -> Imho its usefull that you need to apply functions and then merge results of this functions together. Example: calculating avg score. First argument to converge is function which will 'merge/connect' result of the array of function passed in 2nd argument. R.converge(R.divide, [R.sum, R.length]) . So first step is that you got result from sum and length, and its passed to 'merge' function -> and the end you finish with ONE result

AP is different -> Firstly you will not end with one result, but an array. and as you have mention it just apply functions on a data set.. And its JUST concatenated,
R.ap([R.multiply(2), R.add(3)], [1,2,3]); First step apply R.multiply, and you get [2,4,6], 2nd step [4,5,6], concatenate and you got final result [2,4,6,4,5,6]

Of course you can also write AP, using converage with 'merge' function set to 'contact'

R.converge(R.concat, [R.map(R.multiply(2)), R.map(R.add(3))])([1,2,3])

For me converge is like 'splitting' value to two functions, do some calculations, and merge them back (I've noticed its usefull for point free, if you need to get the argument in two or more places)

Isaac
@IsaacRaja_twitter
Sep 14 2017 06:46
@mimol91 Thanks! It clears up a bit. But then const isPalindrome = R.ap(R.equals, R.reverse)('test') i don't understand how this works
David Sancho
@davesnx
Sep 14 2017 06:46
Hey guys, it’s there a better way (in terms of readability) to ge the minumum of a list based on minBy?
I remember a issue minListBy, but didn’t progress
any ideas?
Adam Szaraniec
@mimol91
Sep 14 2017 06:48
@IsaacRaja_twitter Dont worry for me also
const isPalindrome = s => s === R.reverse(s) is much easier to understand
Adam Szaraniec
@mimol91
Sep 14 2017 07:56
@svozza Could you explain how does this palindrome works? Unfortunatelly ramda doc has all examples with an array as argument
Stefano Vozza
@svozza
Sep 14 2017 08:18
yeah, no problem. if you look in the docs it contains the cryptic sentence Also treats curried functions as applicatives.
what that means in reality is that when you pass two functions to ap rather than arrays they get treated like this:
const h = f => g => x => f(x, g(x));
the tests in ramda are very instructive
so R.ap(R.equals, R.reverse) === x => R.equals(x, R.reverse(x))
Adam Szaraniec
@mimol91
Sep 14 2017 08:28
hmm, my mind is blowing...
is it possible to force converge to works on single value instead of array
How can I make it to be equal?
R.converge(R.add, [R.idenity, R.idenity])(10)  === R.ap(R.add, R.identity)(10)
Stefano Vozza
@svozza
Sep 14 2017 08:33
so like? const palindrome = R.converge(R.equals, [R.identity, R.reverse])
oh sorry, i misunderstood your question
your example works, although you have a typo in R.idenTity
@ram-bot
R.converge(R.add, [R.identity, R.identity])(10)
ram-bot
@ram-bot
Sep 14 2017 08:37
20
Stefano Vozza
@svozza
Sep 14 2017 08:38
@ram-bot
R.ap(R.add, R.identity)(10)
ram-bot
@ram-bot
Sep 14 2017 08:38
20
Adam Szaraniec
@mimol91
Sep 14 2017 09:58
ahh,, thanks
I like, that almost everyday I m learning sth new here =)
Stefano Vozza
@svozza
Sep 14 2017 10:06
actually that's a good way of thinking about it if you're used to using converge. Any time you find yourself writing R.converge(f, [R.identity, g]) you can use ap
hkrutzer
@hkrutzer
Sep 14 2017 11:08
Does anyone use Typescript with Ramda? Is there a way to typecheck the 'thing' field in R.prop('thing', a)?
Brad Compton (he/him)
@Bradcomp
Sep 14 2017 15:47
I think @tycho01 is the person here with the most experience with Ramda and Typescript