Yeah, this thread has been a gold mine for me -- lots to study, to really understand how they all work, but very useful. Also in particular thank you Joe, I think you and I started learning J around the same time and it's a bit of an inspiration to see how quickly you're picking it up. I'll have to step up my game =)
Thanks to everyone for all of the tips and tricks! -Dan On Tue, Oct 8, 2013 at 7:05 AM, Joe Bogner <[email protected]> wrote: > I wrote: >> Is there a better way to find indices of a subset within a greater list? > This is my hacky solution > > Sorry, please disregard my question. I figured out a cleaner implementation > using the table adverb > > (; > L:0 I. each <"1 (find =/ list)) { list > > NB. 13 : '(; > L:0 I. each <"1 (x =/ y)) { y' > in=:] {~ [: ; [: >L:0 [: I.&.> [: <"1 =/ > > ] find in list > > ┌─┬─┬─┬─┐ > > │a│a│b│q│ > > └─┴─┴─┴─┘ > > > ] (<'a') in ('a';'b') > > ┌─┐ > > │a│ > > └─┘ > > > ] ('a';'b') in ('a';'b';'c';'a') > > ┌─┬─┬─┐ > > │a│a│b│ > > └─┴─┴─┘ > > > > idx=:13 : '(; > L:0 I. each <"1 (x =/ y))' > > > Which then lets us do things like the following to merge/join tables > > > (find idx list) { list > > > My previous hack of namesTenRows=.(; L:0 > (3 : '(<y) ([: I. =) names' ) > each namesTen) { data > > Becomes: namesTenRows=. (namesTen idx names) { data > > > Also quite a bit faster than the hacky solution > > > ]1e5 * 1e5 (6!:2) '; > L:0 (3 : ''(<y) ([: I. =) list'' ) each find' > > 0.589178 > > > ]1e5 * 1e5 (6!:2) '(find idx list)' > > 0.224204 > > > Fun fun > > > http://www.jsoftware.com/help/dictionary/d420.htm > http://www.jsoftware.com/help/learning/18.htm > > I'd be happy to contribute to a wiki page on slicing/dicing tables. I don't > have access to modify the wiki at present. > > > On Tue, Oct 8, 2013 at 9:20 AM, Joe Bogner <[email protected]> wrote: > >> One more -- matching to a list (joining tables) >> >> NB. Get the names of everyone from the frequency table with exactly 10 >> matches >> >> ]namesTen=. > 0}"1 (10 ([: I. =) (> 1{"1 nameFrequency)) { nameFrequency >> >> ┌─────┬─────┬─────┬─────┬─────┬──────┬─────┬──────┬────────┬─────┐ >> >> │Ruben│Paddy│Cecil│Chico│Edwin│Martin│Kelly│Junior│Jonathan│Duane│ >> >> └─────┴─────┴─────┴─────┴─────┴──────┴─────┴──────┴────────┴─────┘ >> >> >> >> namesTenRows=.(; L:0 > (3 : '(<y) ([: I. =) names' ) each namesTen) { data >> >> >> NB. Get the unique names in our subset to validate it matches the above >> >> ] ~. (colIdx <'nameFirst') {"1 namesTenRows >> >> ┌─────┬─────┬─────┬─────┬─────┬──────┬─────┬──────┬────────┬─────┐ >> >> │Ruben│Paddy│Cecil│Chico│Edwin│Martin│Kelly│Junior│Jonathan│Duane│ >> >> └─────┴─────┴─────┴─────┴─────┴──────┴─────┴──────┴────────┴─────┘ >> >> >> >> Is there a better way to find indices of a subset within a greater list? >> This is my hacky solution >> >> >> find=.('a';'b';'q') >> >> list=.('a';'b';'z';'c';'a';'q') >> >> >> ] (; > L:0 (3 : '(<y) ([: I. =) list' ) each find) { list >> >> ┌─┬─┬─┬─┐ >> >> │a│a│b│q│ >> >> └─┴─┴─┴─┘ >> >> >> find xxx list >> >> >> >> On Tue, Oct 8, 2013 at 7:14 AM, Ric Sherlock <[email protected]> wrote: >> >>> A relatively simple adverb like the one below can be useful if you a >>> repeatedly trying to access different fields in the table: >>> >>> NB.*getFields a Returns fields (columns) from y specified in x >>> >>> NB. EG: 'fieldname' tablehdr getFields table >>> >>> NB. EG: ('Lastname';'Firstname') hdr getFields dat >>> >>> NB. y is: table of records >>> >>> NB. m is: list of boxed field names in y >>> >>> NB. x is: list of boxed field names to return >>> >>> getFields=: 1 : 0 >>> >>> : >>> >>> flds=. boxopen x >>> >>> idx=. m i. flds >>> >>> idx {"1 y >>> >>> ) >>> >>> >>> require 'tables/csv' >>> >>> 'hdr dat'=: split readcsv 'c:/tmp/master.csv' >>> >>> >>> 4{. dat hdr getFields~ 'nameFirst';'birthYear';'weight' >>> >>> >>> ┌──────┬────┬───┐ >>> >>> │Hank │1934│180│ >>> >>> ├──────┼────┼───┤ >>> >>> │Tommie│1939│190│ >>> >>> ├──────┼────┼───┤ >>> >>> │Don │1954│190│ >>> >>> ├──────┼────┼───┤ >>> >>> │Andy │1972│184│ >>> >>> └──────┴────┴───┘ >>> >>> >>> >>> >>> On Tue, Oct 8, 2013 at 11:39 PM, Joe Bogner <[email protected]> wrote: >>> >>> > Here are a couple other examples: >>> > >>> > NB. Convert years to a flat list of numbers >>> > >>> > birthYears=. , L:0 (> ". each ((colIdx <'birthYear') {"1 data)) >>> > >>> > >>> > NB. Find players born between 1930 and 1950 named Hank >>> > >>> > (I. (birthYears > 1930) * (birthYears < 1950) * (names = <'Hank')) { >>> data >>> > >>> > >>> > NB. Find players born between 1930 and 1950 named Hank or Tommie >>> > >>> > (I. (birthYears > 1930) * (birthYears < 1950) * ( (names = <'Hank') + >>> > (names = <'Tommie'))) { data >>> > >>> > >>> > >>> > NB. http://www.jsoftware.com/jwiki/Essays/DataStructures >>> > >>> > diff=: -. [ >>> > >>> > intersect=: e. # [ >>> > >>> > union=: ~.@, >>> > >>> > >>> > in=: ] ([: I. =)~ [: < [ NB. 4 : '(<x) ([: I. =) >>> y' >>> > >>> > inT=: 4 : 'x in (colIdx <y) {"1 data' >>> > >>> > >>> > NB. Another style >>> > >>> > NB. Hank or Tommie - the union of indices >>> > >>> > (('Hank' inT 'nameFirst') union ('Tommie' inT 'nameFirst')) { data >>> > >>> > >>> > NB. Hank or Tommie AND born in Cuba >>> > >>> > ((('Hank' inT 'nameFirst') union ('Tommie' inT 'nameFirst')) intersect >>> > ('Cuba' inT 'birthCountry')) { data >>> > >>> > >>> > >>> > On Tue, Oct 8, 2013 at 3:50 AM, R.E. Boss <[email protected]> wrote: >>> > >>> > > So 'data' was plural? >>> > > >>> > > >>> > > R.E. Boss >>> > > >>> > > (Add your info to >>> http://www.jsoftware.com/jwiki/Community/Demographics) >>> > > >>> > > >>> > > > -----Original Message----- >>> > > > From: [email protected] [mailto:programming- >>> > > > [email protected]] On Behalf Of Henry Rich >>> > > > Sent: dinsdag 8 oktober 2013 3:52 >>> > > > To: [email protected] >>> > > > Subject: Re: [Jprogramming] Slicing and dicing tables >>> > > > >>> > > > That is what he meant, but it's a minority opinion. >>> > > > >>> > > > Henry Rich >>> > > > >>> > > > On 10/7/2013 8:56 PM, Dan Bron wrote: >>> > > > > You mean " 'data' is plural " . >>> > > > > >>> > > > > -Dan >>> > > > > >>> > > > > PS: http://en.m.wikipedia.org/wiki/Muphry's_law >>> > > > > >>> > > > > (Yes, that link is spelled correctly :) >>> > > > > >>> > > > > Please excuse typos; composed on a handheld device. >>> > > > > >>> > > > > On Oct 7, 2013, at 8:00 PM, Eldon Eller <[email protected]> >>> wrote: >>> > > > > >>> > > > >> Data are plural. >>> > > > >> >>> > > > >> On 10/07/2013 02:04 PM, Dan Bron wrote: >>> > > > >>> Ganesh Rapolu >>> > > > >>>> Because the data is boxed, all comparisons must be boxed. >>> > > > >>> This was very well put (one short sentence which both clarifies >>> the >>> > > > >>> problem and justifies the solution). >>> > > > >>> >>> > > > >>> -Dan >>> > > > >>> >>> > > > >>> >>> > > > >>> >>> > > > >>> >>> > -------------------------------------------------------------------- >>> > > > >>> -- For information about J forums see >>> > > > >>> http://www.jsoftware.com/forums.htm >>> > > > >> >>> > > > >> >>> > --------------------------------------------------------------------- >>> > > > >> - For information about J forums see >>> > > > >> http://www.jsoftware.com/forums.htm >>> > > > > >>> > ---------------------------------------------------------------------- >>> > > > > For information about J forums see >>> > > > http://www.jsoftware.com/forums.htm >>> > > > > >>> > > > >>> ---------------------------------------------------------------------- >>> > > > For information about J forums see >>> http://www.jsoftware.com/forums.htm >>> > > >>> > > ---------------------------------------------------------------------- >>> > > For information about J forums see >>> http://www.jsoftware.com/forums.htm >>> > > >>> > ---------------------------------------------------------------------- >>> > For information about J forums see http://www.jsoftware.com/forums.htm >>> > >>> ---------------------------------------------------------------------- >>> For information about J forums see http://www.jsoftware.com/forums.htm >>> >> >> > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
