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

Reply via email to