> What I'm looking for is a natural, conceptually clean approach.
Erlang programmers do version 2 all the time and it can be call
"Erlang-style". You should be confident with version 2.
Version 2 is cleaner if you write the inner loop as a separate
function.
--
You received this message because
Hi,
On Aug 9, 8:06 am, gary ng wrote:
> Assuming vector is implemented in some form of array(well it seems to
> have the characteristic of an array where accessing any element would
> be O(1)), appending one element can be O(1) if there are space
> reserved for some extra items but if that actio
On Sun, Aug 8, 2010 at 10:50 PM, Meikel Brandmeyer wrote:
> Hi,
>
> On Aug 8, 6:19 pm, gary ng wrote:
>
>> I wrote a similar version in F# which does have the advantage of
>> handling infinite input or a very long partition in the sense that I
>> can still consume the first element immediately an
On Sun, Aug 8, 2010 at 10:32 PM, Meikel Brandmeyer wrote:
> I'm bit confused about what you mean here, but vector append (read:
> conj) is O(1) (don't nail me down on whether it's amortised). There is
> no array copying going on underneath.
>
Assuming vector is implemented in some form of array(we
Hi,
On Aug 8, 6:19 pm, gary ng wrote:
> I wrote a similar version in F# which does have the advantage of
> handling infinite input or a very long partition in the sense that I
> can still consume the first element immediately and can skip to the
> second, third group ... In a sense, the result i
Hi,
On Aug 8, 8:25 pm, gary ng wrote:
> Interesting to know. Just curious which one is more efficient(not that
> I question the vector append is slow) as it would still involve some
> array copy and creation(though I assume it would be very efficient as
> it is using the very low level i.e. sort
On Sat, Aug 7, 2010 at 9:30 PM, David Cabana wrote:
> conj is not the same as append; it will insert the new element in the
> smart (most efficient) way. For instance:
> user> (conj '(1 2 3) 0)
> (0 1 2 3)
> user> (conj [1 2 3] 0)
> [1 2 3 0]
>
> There is no performance hit from using conj to in
On Sun, Aug 8, 2010 at 12:12 AM, Meikel Brandmeyer wrote:
> Just for fun another low-level solution:
>
> (defn partition-when
> [pred coll]
> (let [step (fn [p s]
> (if s
> (let [fst (first s)]
> (if (pred fst)
> [p s]
>
Here's my take using my clj-iter (http://github.com/nathell/clj-iter):
(defn partition-when
[pred coll]
(iter (for x in coll)
(for p = (when x (pred x)))
(for y initially () then (if p (list x) (cons x y)))
(collect (reverse y) if p)
(finally-collect (reverse y)
Hi,
Am 08.08.2010 um 05:27 schrieb David Cabana:
> Using a vector instead of a list as the accumulator makes it possible
> to skip the mapping of reverse used in the earlier version of pw:
>
> (defn pw [f? x]
> (let [phi (fn [a e]
> (if (f? e)
>(cons [e] a )
>
> conj sounds like 'append' to me which I have no idea about the
> performance characteristics in clojure(it is a no-no in F#, Haskell ++
> is better but would grow the stack).
conj is not the same as append; it will insert the new element in the
smart (most efficient) way. For instance:
user> (co
On Sat, Aug 7, 2010 at 8:27 PM, David Cabana wrote:
> Using a vector instead of a list as the accumulator makes it possible
> to skip the mapping of reverse used in the earlier version of pw:
>
> (defn pw [f? x]
> (let [phi (fn [a e]
> (if (f? e)
> (cons [e] a )
>
Using a vector instead of a list as the accumulator makes it possible
to skip the mapping of reverse used in the earlier version of pw:
(defn pw [f? x]
(let [phi (fn [a e]
(if (f? e)
(cons [e] a )
(cons (conj (first a) e)
(rest
On Sat, Aug 7, 2010 at 9:36 PM, gary ng wrote:
> if you don't mind about performance, this seems to be natural to me
>
> user=> (reverse (map reverse (reduce (fn [a e] (if (even? e) (cons [e] a)
> (cons
> (cons e (first a)) (rest a (list) [1 2 3 7 5 4 1])))
> ((1) (2 3 7 5) (4 1))
I reworked
On Sat, Aug 7, 2010 at 6:55 PM, Michael Gardner wrote:
> On Aug 7, 2010, at 8:48 PM, Michael Gardner wrote:
>
>> On Aug 7, 2010, at 8:39 PM, gary ng wrote:
>>
>>> nice, why do I need the gensym ?
>>
>> The gensym was just a cheesy way of generating a unique value.
>
> To elaborate a bit more on my
On Aug 7, 2010, at 8:48 PM, Michael Gardner wrote:
> On Aug 7, 2010, at 8:39 PM, gary ng wrote:
>
>> nice, why do I need the gensym ?
>
> The gensym was just a cheesy way of generating a unique value.
To elaborate a bit more on my failure, I was reading the docs for partition-by
thinking that
On Aug 7, 2010, at 8:35 PM, Michael Gardner wrote:
> (partition-by #(when (even? %) (gensym))
Whoops, hit 'send' too soon. And it doesn't actually work, since it splits
before and after the even values.
On Aug 7, 2010, at 8:39 PM, gary ng wrote:
> nice, why do I need the gensym ?
The gensym w
On Sat, Aug 7, 2010 at 6:39 PM, gary ng wrote:
> On Sat, Aug 7, 2010 at 6:35 PM, Michael Gardner wrote:
>> (partition-by #(when (even? %) (gensym))
> nice, why do I need the gensym ?
ah, the gensym is for 'break'. but why does it behave like this
user=> (partition-by #(when (even? %) (gensym))
On Sat, Aug 7, 2010 at 6:35 PM, Michael Gardner wrote:
> (partition-by #(when (even? %) (gensym))
nice, why do I need the gensym ?
--
You received this message because you are subscribed to the Google
Groups "Clojure" group.
To post to this group, send email to clojure@googlegroups.com
Note that
On Sat, Aug 7, 2010 at 6:12 PM, David Cabana wrote:
> Speed is no big deal for me; the sequences I'm handling are short, say
> about length100. BTW, one of these is considerably faster than the
> other for longish sequences. Can you guess which?
>
if you don't mind about performance, this seems to
(partition-by #(when (even? %) (gensym))
--
You received this message because you are subscribed to the Google
Groups "Clojure" group.
To post to this group, send email to clojure@googlegroups.com
Note that posts from new members are moderated - please be patient with your
first post.
To unsubsc
Here are a couple of implementations of a function I'm calling
'partition-when'. I feel like there should be a simpler way than
either of these. If you have one, I'd love to see it.
(defn partition-when;;version 1
"Partition a sequence into subsequences; begin a new
s
22 matches
Mail list logo
