Function Design: sequence or "&" argument?
If you have a function that needs to treat multiple arguments as a group, what forces drive you to represent this as a single sequence argument vs. an "&" argument? To give a concrete example, why does "+" work like (+ 1 2 3 4) instead of (+ [1 2 3 4]) Is it performance? Aesthetics? Composability concerns? Not having to call "apply" all the time? Thanks, Chris -- 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 unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Function Design: sequence or "&" argument?
On Mon, Nov 15, 2010 at 11:52 AM, Chris wrote: > If you have a function that needs to treat multiple arguments as a > group, what forces drive you to represent this as a single sequence > argument vs. an "&" argument? To give a concrete example, why does > "+" work like > > (+ 1 2 3 4) > > instead of > > (+ [1 2 3 4]) > > Is it performance? Aesthetics? Composability concerns? Not having > to call "apply" all the time? "Not having to use extra brackets all the time" gets my vote. Yeah, yeah, I know, a strange statement coming from a lisp-head. -- 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 unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Function Design: sequence or "&" argument?
Hi, Am 15.11.2010 um 17:52 schrieb Chris: > If you have a function that needs to treat multiple arguments as a > group, what forces drive you to represent this as a single sequence > argument vs. an "&" argument? To give a concrete example, why does > "+" work like > > (+ 1 2 3 4) > > instead of > > (+ [1 2 3 4]) > > Is it performance? Aesthetics? Composability concerns? Not having > to call "apply" all the time? Semantics? + is a binary operator. Allowing more arguments is a convenience I guess. The mathematically honest definition would have arity 2. Eg. filter on the other hand acts on a sequence. That's why it is not defined as (filter even? 1 2 3 4). (See also map for example where apply wouldn't help.) Does this make sense? Sincerely Meikel -- 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 unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Function Design: sequence or "&" argument?
Performance is part of it too. Allowing dispatch on arity leads to faster code. Many of the functions that operate on sequences are lazy so dispatch on arity doesn't apply. On Nov 15, 11:52 am, Chris wrote: > If you have a function that needs to treat multiple arguments as a > group, what forces drive you to represent this as a single sequence > argument vs. an "&" argument? To give a concrete example, why does > "+" work like > > (+ 1 2 3 4) > > instead of > > (+ [1 2 3 4]) > > Is it performance? Aesthetics? Composability concerns? Not having > to call "apply" all the time? > > Thanks, > Chris -- 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 unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Function Design: sequence or "&" argument?
That makes sense... thanks, Meikel. Maybe my example of + wasn't the best, given its mathematical nature. Here's my situation: I'm writing some software to analyze some protein datasets, part of which entails generating a Venn diagram of their intersections. Each dataset has a unique name, and my function looks up the data by this name in a database. Right now the function takes a single "&" parameter of "datasets"; you give it as many dataset names as you want, and it gives you the intersection breakdown. Internally, this function just treats "datasets" like the sequence it is. So I guess a better way to state my question would be this: if you only have a function that operates on a sequence of items, and it could be written either as (def foo [& stuff] ... ) or (def foo [stuff] ... ) ;; where 'stuff' is a sequence are there any concerns besides (for lack of a better term) the "ergonomics" of calling such functions that should influence which signature to use? Does that make sense? Thanks, Chris On Mon, Nov 15, 2010 at 2:08 PM, Meikel Brandmeyer wrote: > Hi, > > Am 15.11.2010 um 17:52 schrieb Chris: > >> If you have a function that needs to treat multiple arguments as a >> group, what forces drive you to represent this as a single sequence >> argument vs. an "&" argument? To give a concrete example, why does >> "+" work like >> >> (+ 1 2 3 4) >> >> instead of >> >> (+ [1 2 3 4]) >> >> Is it performance? Aesthetics? Composability concerns? Not having >> to call "apply" all the time? > > Semantics? > > + is a binary operator. Allowing more arguments is a convenience I guess. The > mathematically honest definition would have arity 2. > > Eg. filter on the other hand acts on a sequence. That's why it is not defined > as (filter even? 1 2 3 4). (See also map for example where apply wouldn't > help.) > > Does this make sense? > > Sincerely > Meikel > > -- > 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 unsubscribe from this group, send email to > clojure+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/clojure?hl=en -- 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 unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Function Design: sequence or "&" argument?
Hi, On Tue, Nov 16, 2010 at 5:21 PM, Chris Maier wrote: > That makes sense... thanks, Meikel. > > Maybe my example of + wasn't the best, given its mathematical nature. > > Here's my situation: I'm writing some software to analyze some protein > datasets, part of which entails generating a Venn diagram of their > intersections. Each dataset has a unique name, and my function looks > up the data by this name in a database. Right now the function takes > a single "&" parameter of "datasets"; you give it as many dataset > names as you want, and it gives you the intersection breakdown. > Internally, this function just treats "datasets" like the sequence it > is. > > So I guess a better way to state my question would be this: if you > only have a function that operates on a sequence of items, and it > could be written either as > > (def foo [& stuff] ... ) > > or > > (def foo [stuff] ... ) ;; where 'stuff' is a sequence > > are there any concerns besides (for lack of a better term) the > "ergonomics" of calling such functions that should influence which > signature to use? > If you really want to treat your sequence lazily pass it explicitely and not as varargs. You can also provide both: clojure.core has #'conj and #'into for example. hth, Christophe -- 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 unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Function Design: sequence or "&" argument?
Ah, laziness... thanks Christophe. For my particular application laziness doesn't matter (so that hadn't occurred to me) but that's a good general principle to keep in mind. Thanks, Chris On Tue, Nov 16, 2010 at 11:31 AM, Christophe Grand wrote: > Hi, > > On Tue, Nov 16, 2010 at 5:21 PM, Chris Maier > wrote: >> >> That makes sense... thanks, Meikel. >> >> Maybe my example of + wasn't the best, given its mathematical nature. >> >> Here's my situation: I'm writing some software to analyze some protein >> datasets, part of which entails generating a Venn diagram of their >> intersections. Each dataset has a unique name, and my function looks >> up the data by this name in a database. Right now the function takes >> a single "&" parameter of "datasets"; you give it as many dataset >> names as you want, and it gives you the intersection breakdown. >> Internally, this function just treats "datasets" like the sequence it >> is. >> >> So I guess a better way to state my question would be this: if you >> only have a function that operates on a sequence of items, and it >> could be written either as >> >> (def foo [& stuff] ... ) >> >> or >> >> (def foo [stuff] ... ) ;; where 'stuff' is a sequence >> >> are there any concerns besides (for lack of a better term) the >> "ergonomics" of calling such functions that should influence which >> signature to use? > > If you really want to treat your sequence lazily pass it explicitely and not > as varargs. > > You can also provide both: clojure.core has #'conj and #'into for example. > > hth, > > Christophe > > -- > 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 unsubscribe from this group, send email to > clojure+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/clojure?hl=en -- 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 unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Function Design: sequence or "&" argument?
On Nov 15, 8:52 am, Chris wrote: > If you have a function that needs to treat multiple arguments as a > group, what forces drive you to represent this as a single sequence > argument vs. an "&" argument? To give a concrete example, why does > "+" work like > > (+ 1 2 3 4) > > instead of > > (+ [1 2 3 4]) > > Is it performance? Aesthetics? Composability concerns? Not having > to call "apply" all the time? > > Thanks, > Chris The "&" form is a syntax helper. In other words it is intended to change the source code, not to change the way you program. Thus you have to look at the intended use of the function. If you are going to type different number arguments to a function right in the code yourself, then use &. If you are going to get the arguments for that function at runtime, then use a single argument. Examples: map, filter, reduce. -- 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 unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Function Design: sequence or "&" argument?
I always write a function to take varargs because it can also take a list using apply. (+ 1 2 3 4 5) (apply + [1 2 3 4 5]) On Nov 15, 9:52 am, Chris wrote: > If you have a function that needs to treat multiple arguments as a > group, what forces drive you to represent this as a single sequence > argument vs. an "&" argument? To give a concrete example, why does > "+" work like > > (+ 1 2 3 4) > > instead of > > (+ [1 2 3 4]) > > Is it performance? Aesthetics? Composability concerns? Not having > to call "apply" all the time? > > Thanks, > Chris -- 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 unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Function Design: sequence or "&" argument?
I always write a function to take a single seq argument because it can also take varargs if I wrap them in a seq. (defn add [nums] (reduce + nums)) (add some-seq) (add [1 2 3 4 5]) On Nov 19, 4:19 pm, Jarl Haggerty wrote: > I always write a function to take varargs because it can also take a > list using apply. > > (+ 1 2 3 4 5) > (apply + [1 2 3 4 5]) > > On Nov 15, 9:52 am, Chris wrote: > > > If you have a function that needs to treat multiple arguments as a > > group, what forces drive you to represent this as a single sequence > > argument vs. an "&" argument? To give a concrete example, why does > > "+" work like > > > (+ 1 2 3 4) > > > instead of > > > (+ [1 2 3 4]) > > > Is it performance? Aesthetics? Composability concerns? Not having > > to call "apply" all the time? > > > Thanks, > > Chris > > -- 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 unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Function Design: sequence or "&" argument?
I generally prefer to pass in a sequence rather than use a variable number of arguments. The only time variable arguments are really useful is in functions like map (or maybe +) in which you rarely use more than one (or two) arguments and it would be a pain to wrap the last argument in a list. e.g. #+begin_src clojure (map (partial + 1) (list (range 10))) #+end_src would be very awkward but required without &rest finally, explicitly wrapped sequences are more composable just my opinions... -- Eric Alan writes: > I always write a function to take a single seq argument because it can > also take varargs if I wrap them in a seq. > > (defn add [nums] > (reduce + nums)) > > (add some-seq) > (add [1 2 3 4 5]) > > On Nov 19, 4:19 pm, Jarl Haggerty wrote: >> I always write a function to take varargs because it can also take a >> list using apply. >> >> (+ 1 2 3 4 5) >> (apply + [1 2 3 4 5]) >> >> On Nov 15, 9:52 am, Chris wrote: >> >> > If you have a function that needs to treat multiple arguments as a >> > group, what forces drive you to represent this as a single sequence >> > argument vs. an "&" argument? To give a concrete example, why does >> > "+" work like >> >> > (+ 1 2 3 4) >> >> > instead of >> >> > (+ [1 2 3 4]) >> >> > Is it performance? Aesthetics? Composability concerns? Not having >> > to call "apply" all the time? >> >> > Thanks, >> > Chris >> >> -- 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 unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
