c=: 'comb4=: [ (([ = [: +/"1 ]) ([: |. [: I. #) ]) [: #: [: i. 2 ^ ]'
+/\ 1 _1 0{~ '()' i. c
0 0 0 0 0 0 0 0 0 0 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 2 2 2 2 2 2 2 2 2 2 2
2 2 2 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
c,:1":+/\ 1 _1 0{~ '()' i. c
comb4=: [ (([ = [: +/"1 ]) ([: |. [: I. #) ]) [: #: [: i. 2 ^ ]
000000000012222222222222211222222222222221110000000000000000000
The parens look matched to me.
On Wed, Jun 12, 2013 at 7:59 AM, Linda Alvord <[email protected]>wrote:
> Ric, Look closely at the display of comb4 (whichav works). Have you
> cever seen a function with unmatched parentheses?
>
> comb4=: [ (([ = [: +/"1 ]) ([: |. [: I. #) ]) [: #: [: i. 2 ^ ]
> 2 comb4 4
> 0 1
> 0 2
> 0 3
> 1 2
> 1 3
> 2 3
>
> comb4
> [ (([ = [: +/"1 ]) ([: |. [: I. #) ]) [: #: [: i. 2 ^ ]
>
>
> Now look at the definition I want to write:
>
> comb5=: 13 :'((x=+/"1 y) (|.I.#) #:i.2^y'
>
> comb5
> 4 : '((x=+/"1 y) (|.I.#) #:i.2^y'
>
> 2 comb5 4
> |syntax error: comb5
> | ((x=+/"1 y)(|.I.#)#:i.2^y
>
>
> My explicit comb5 is the one I want to work. It shows the structure of
> combinations beautifully:
>
>
> x (f g h) y evaluates as (x f y) g (x h y)
>
> Linda
>
>
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of Linda
> Alvord
> Sent: Wednesday, June 12, 2013 10:20 AM
> To: [email protected]
> Subject: Re: [Jprogramming] z458095869_dissectnopath_
>
> Ric, comb3 works, but when you display the definition it inserts a set of
> parentheses that are necessary to determine the order of operation. Can
> you
> make your version work with doing that?
>
> comb2=: 13 :'|. I. (x=+/"1 #:i.2^y)##:i.2^y'
>
> comb3=: ((= +/"1) ([: |. [: I. #) ]) [: #: [: i. 2 ^ ]
> 2 comb3 4
> 0 1
> 0 2
> 0 3
> 1 2
> 1 3
> 2 3
>
> comb2
> [: |. [: I. ([ = [: +/"1 [: #: [: i. 2 ^ ]) # [: #: [: i. 2 ^ ]
>
> comb3
> ((= +/"1) ([: |. [: I. #) ]) ([: #: [: i. 2 ^ ])
>
> I haven't worked on comb4 yet, but I'm looking forward to it.
>
> Linda
>
>
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of Ric
> Sherlock
> Sent: Tuesday, June 11, 2013 6:31 AM
> To: Programming JForum
> Subject: Re: [Jprogramming] z458095869_dissectnopath_
>
> Linda,
> In case it is of any interest to you, the following is a more faithful
> translation of the original comb1 to a tacit form without @ @: or &
> comb3=: ((= +/"1) ([: |. [: I. #) ]) ([: #: [: i. 2 ^ ])
>
> If you also want to remove hooks then ...
> comb4=: [ (([ = [: +/"1 ]) ([: |. [: I. #) ]) [: #: [: i. 2 ^ ]
>
> Have you tried creating tacit forms without using 13 : ? If not I'd
> encourage you to give it a go!
>
>
> On Tue, Jun 11, 2013 at 7:56 PM, Linda Alvord
> <[email protected]>wrote:
>
> > It took quite a while to remove @ @: & from Roger's great
> > combinations function and I learned a lot. You may find these
> comparisons fun.
> >
> > If you try this script in J801 jhs then use NB. to ignore the two ds
> > statements
> >
> >
> > require '~addons/debug/dissect/dissect.ijs'
> >
> > comb1=: ((= +/"1) |.@:I.@# ]) #:@i.@(2&^)
> > 2 comb1 4
> > comb1
> >
> > comb2=: 13 :'|. I. (x=+/"1 #:i.2^y)##:i.2^y'
> > 2 comb2 4
> > comb2
> >
> > ds '2([: |. [: I. ([ = [: +/"1 [: #: [: i. 2 ^ ]) # [: #: [: i. 2 ^ ])4'
> >
> > ds '2(((= +/"1) |.@:I.@# ]) #:@i.@(2&^))4'
> >
> > 5!:4 <'comb1'
> >
> > 5!:4 <'comb2'
> >
> > Linda
> >
> >
> > -----Original Message-----
> > From: [email protected]
> > [mailto:[email protected]] On Behalf Of Linda
> > Alvord
> > Sent: Friday, June 07, 2013 3:55 PM
> > To: [email protected]
> > Subject: Re: [Jprogramming] z458095869_dissectnopath_
> >
> > Here's another happy story about frequency distributions.
> >
> > require '~addons/debug/dissect/dissect.ijs'
> >
> > ]A=:?4 5$6
> >
> > fd=: 13 :'/:~({.,#)/.~,y'
> > fd A
> >
> > fd2=: 13 :'/:~ (,y)({.,#)/.,y'
> > fd2 A
> >
> > fd
> > fd2
> >
> > ds '([: /:~ , ({. , #)/. ,)?4 5$6'
> > ds '([: /:~ [: ({. , #)/.~ ,)?4 5$6'
> >
> > Enter the script above. I never understood fd until today. Looking
> > at your ds version allowed me to write an explicit version. The
> > tacit result was even shorter! It continues to be a kind of miracle
> > tool.
> >
> > The newer ds is an improvement too.
> >
> > Linda
> >
> > -----Original Message-----
> > From: [email protected]
> > [mailto:[email protected]] On Behalf Of Henry
> > Rich
> > Sent: Thursday, June 06, 2013 10:12 AM
> > To: [email protected]
> > Subject: Re: [Jprogramming] z458095869_dissectnopath_
> >
> > You can use a function in a function; ds just won't look inside the name.
> >
> > Henry Rich
> >
> > On 6/6/2013 2:02 AM, Linda Alvord wrote:
> > > Henry, Earlier I thought I couldn't use a function inside another
> > > function, but it seems now that I can.
> > >
> > > require '~addons/debug/dissect/dissect.ijs'
> > >
> > > pas=: 13 :'(i.>:y)!y'
> > >
> > > NB. BINOMIAL PROBABILITIES
> > > bp=: 13 :' (pas x)*(y^ i.>:x)*(1-y)^ |. i.>:x'
> > >
> > > ds 'pas 3'
> > > ds '3(([: pas [) * (] ^ [: i. [: >: [) * (1 - ]) ^ [: |. [: i. [: >:
> > [)0.9'
> > >
> > > Linda
> > >
> > > -----Original Message-----
> > > From: [email protected]
> > > [mailto:[email protected]] On Behalf Of Linda
> > > Alvord
> > > Sent: Wednesday, June 05, 2013 9:54 PM
> > > To: [email protected]
> > > Subject: Re: [Jprogramming] z458095869_dissectnopath_
> > >
> > > I just realized what I did. By mistake I wrote ds=: and realized
> > > it and knew I had to restart jqt. When I did all was fine.
> > >
> > > Sorry to bother you.
> > >
> > > Linda
> > >
> > >
> > > -----Original Message-----d
> > > Froam: [email protected]
> > > [mailto:[email protected]] On Behalf Of Henry
> > > Rich
> > > Sent: Wednesday, June 05, 2013 9:26 PM
> > > To: [email protected]
> > > Subject: Re: [Jprogramming] z458095869_dissectnopath_
> > >
> > > Are you sure? When I run
> > >
> > > ds '2([: |: ([ = [: +/ [: |: #~ #: [: i. ^) #"1 [: |: #~ #: [: i. ^)4'
> > >
> > > I get a valid display. There seems to be a weird () following the 2
> > > in your version.
> > >
> > > Henry Rich
> > >
> > > On 6/5/2013 9:19 PM, Linda Alvord wrote:
> > >> Henry, I used your ds version of cr and used it to rewrite a
> > >> shorter version of cr.
> > >>
> > >> require '~addons/debug/dissect/dissect.ijs'
> > >>
> > >> cr=: 13 :'|:(x= +/ |:(y#x)#:i.x^y) #"1 |:(y#x)#:i.x^y'
> > >> cr
> > >> [: |: ([ = [: +/ [: |: #~ #: [: i. ^) #"1 [: |: #~ #: [: i. ^
> > >>
> > >> 2 cr 4
> > >> 0 0 1 1
> > >> 0 1 0 1
> > >> 0 1 1 0
> > >> 1 0 0 1
> > >> 1 0 1 0
> > >> 1 1 0 0
> > >>
> > >> ds '2([: |: ([ = [: +/ [: |: #~ #: [: i. ^) #"1 [: |: #~ #: [: i.
> > ^)4'
> > >> |syntax error: ds
> > >> | 2()[:|:([=[:+/[:|:#~#:[:i.^)#"1[:|:#~#:[:i.^)4
> > >>
> > >> Something in the new version is not working.
> > >>
> > >> Linda
> > >>
> > >> -----Original Message-----
> > >> From: [email protected]
> > >> [mailto:[email protected]] On Behalf Of
> > >> Linda Alvord
> > >> Sent: Wednesday, June 05, 2013 7:59 PM
> > >> To: [email protected]
> > >> Subject: Re: [Jprogramming] z458095869_dissectnopath_
> > >>
> > >> Henry, After an upgrade a click on an array in a red box will
> > >> appear in a grid. Now hist 8 appears in the grid with a scroll bar.
> > > Impressive.
> > >>
> > >> Also, 2 cr 5
> > >>
> > >> Thanks. I'm already finding ds useful. Linda
> > >>
> > >>
> > >> -----Original Message-----
> > >> From: [email protected]
> > >> [mailto:[email protected]] On Behalf Of
> > >> Henry Rich
> > >> Sent: Wednesday, June 05, 2013 11:57 AM
> > >> To: [email protected]
> > >> Subject: Re: [Jprogramming] z458095869_dissectnopath_
> > >>
> > >> I have fixed the errors you found, in cr and in launching a grid.
> > >>
> > >> Henry Rich
> > >>
> > >> On 6/5/2013 3:04 AM, Linda Alvord wrote:
> > >>> I'm working on simplifying cr but here is what happens now.
> > >>>
> > >>> cr=:[: |: ([ = [: +/ [: ([: |: ] #: [: i. */) 2 $~ ]) #"1 [:
> ([:
> > >>> |: ]
> > >> #:
> > >>> [: i. */) 2 $~ ]
> > >>> 2 cr 4
> > >>> 0 0 1 1
> > >>> 0 1 0 1
> > >>> 0 1 1 0
> > >>> 1 0 0 1
> > >>> 1 0 1 0
> > >>> 1 1 0 0
> > >>> ds '2 ([: |: ([ = [: +/ [: ([: |: ] #: [: i. */) 2 $~ ]) #"1 [:
> > >>> ([: |: ]
> > >>> #: [: i. */) 2 $~ ])4'
> > >>> length error
> > >>> |length error: combineheights
> > >>> | estheights=:estheights__vop combineheights estheights__uop
> > >>>
> > >>> unknown error during parsing
> > >>>
> > >>> Probably you will spot right away why ds is unhappy.
> > >>>
> > >>> Linda
> > >>> -----Original Message-----
> > >>> From: [email protected]
> > >>> [mailto:[email protected]] On Behalf Of
> > >>> Linda Alvord
> > >>> Sent: Wednesday, June 05, 2013 2:16 AM
> > >>> To: [email protected]
> > >>> Subject: Re: [Jprogramming] z458095869_dissectnopath_
> > >>>
> > >>> Here's a shorter route to the same problem:
> > >>>
> > >>> require '~addons/debug/dissect/dissect.ijs'
> > >>>
> > >>> hist=: 13 :''' O''{~((i.>:y)!y)>/i.>./(i.>:y)!y'
> > >>> hist
> > >>>
> > >>> hist 8
> > >>>
> > >>> ds '('' O'' {~ (] !~ [: i. >:) >/ [: i. [: >./ ] !~ [: i. >:)8'
> > >>>
> > >>> Linda
> > >>>
> > >>> -----Original Message-----
> > >>> From: [email protected]
> > >>> [mailto:[email protected]] On Behalf Of
> > >>> Linda Alvord
> > >>> Sent: Wednesday, June 05, 2013 1:36 AM
> > >>> To: [email protected]
> > >>> Subject: Re: [Jprogramming] z458095869_dissectnopath_
> > >>>
> > >>> Henry,
> > >>>
> > >>> I get an error box if I click on the data in an error box in red:
> > >>>
> > >>> ---------------------------
> > >>> wdhandler
> > >>> ---------------------------
> > >>> error in: dissect_dissectisi_mbldown
> > >>>
> > >>> xywh : command not found: wd
> > >>> (wd ::(''"_)'qer') (13!:8)3
> > >>> ---------------------------
> > >>> OK
> > >>>
> > >>> If I look in the Term window this message is repeated quite a few
> > times.
> > >>>
> > >>>
> > >>> error in paint
> > >>> assertion failure
> > >>> |assertion failure: initevents
> > >>> | 0~:GRIDHWNDC
> > >>>
> > >>> Clicking on data in red boxes doesn't work
> > >>>
> > >>> ---------------------------
> > >>> wdhandler
> > >>> ---------------------------
> > >>> error in: dissect_dissectisi_mbldown
> > >>>
> > >>> xywh : command not found: wd
> > >>> (wd ::(''"_)'qer') (13!:8)3
> > >>> ---------------------------
> > >>> OK
> > >>>
> > >>>
> > >>> error in paint
> > >>> assertion failure
> > >>> |assertion failure: initevents
> > >>> | 0~:GRIDHWNDC
> > >>>
> > >>>
> > >>> JVERSION
> > >>> Engine: j701/2011-01-10/11:25
> > >>> Library: 8.01.011
> > >>> Qt IDE: 1.0.10
> > >>> Platform: Win 32
> > >>> Installer: j801 beta install
> > >>> InstallPath: c:/users/owner/j801
> > >>>
> > >>> All files are upgraded and installed.
> > >>>
> > >>> Maybe this helps.
> > >>>
> > >>> Linda
> > >>>
> > >>> -----Original Message-----
> > >>> From: [email protected]
> > >>> [mailto:[email protected]] On Behalf Of
> > >>> Henry Rich
> > >>> Sent: Tuesday, June 04, 2013 9:43 PM
> > >>> To: [email protected]
> > >>> Subject: Re: [Jprogramming] z458095869_dissectnopath_
> > >>>
> > >>> The result's not too large. It's just not all displayed. The red
> > >>> border around the value is the clue that the value has been
> > >>> truncated for
> > >> display.
> > >>> When you see that, you can click on the truncated value to get a
> > >>> grid that will show you the entire value.
> > >>>
> > >>> Henry Rich
> > >>>
> > >>> On 6/4/2013 9:41 PM, Linda Alvord wrote:
> > >>>> Henry, What I meant was the error that begins when ds indicates
> > >>>> a list of length 70 and only shows a 27 element list. Then
> > >>>> the final table claims shape 9 70 but only shows 9 27.
> > >>>>
> > >>>> Shouldn't there be a message like "result too large" or something?
> > >>>>
> > >>>> Linda
> > >>>>
> > >>>> -----Original Message-----
> > >>>> [mailto:[email protected]]
> > >>>> On Behalf Of Henry Rich
> > >>>> Sent: Tuesday, June 04, 2013 2:53 PM
> > >>>> To: [email protected]
> > >>>> Subject: Re: [Jprogramming] z458095869_dissectnopath_
> > >>>>
> > >>>> OK, I have fixed the grid display. When you have an oversize
> > >>>> result, clicking on it launches a grid to display the entire result.
> > >>>>
> > >>>> Linda, is this what you meant by 'beyond the capacity of ds'?
> > >>>>
> > >>>> Henry Rich
> > >>>>
> > >>>> On 6/4/2013 12:00 AM, Raul Miller wrote:
> > >>>>> Perhaps ds should have scrollbars for displaying large things?
> > >>>>> (Though ideally this should go in wd's implementation, at least
> > >>>>> as an option, and screen resources for things not in view could
> > >>>>> perhaps be avoided. I'm thinking of something like the css
> > >>>>> overflow: scroll mechanism for html.)
> > >>>>>
> > >>>> -----------------------------------------------------------------
> > >>>> --
> > >>>> -
> > >>>> -
> > >>>> - 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
> > >>
> > > --------------------------------------------------------------------
> > > -- 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
>
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
