Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Re: st: no room to add more double literals


From   Sergiy Radyakin <[email protected]>
To   [email protected]
Subject   Re: Re: st: no room to add more double literals
Date   Tue, 18 Jan 2011 11:31:20 -0500

On Tue, Jan 18, 2011 at 11:02 AM, Nick Cox <[email protected]> wrote:
> In response to
>
>> 1) if I try to input a matrix with 1,1,1,1\2,2,2,2\3,3,3,3.... it hits
>> the limit as well with a different error
>> message, but still pretty soon to be practical.
>
> I may be wrenching this out of context, but I note that there is no
> obvious need to spell out every value of a matrix with such simple
> structure. Consider
>
> : J(1,4,(1..20)')
>         1    2    3    4
>     +---------------------+
>   1 |   1    1    1    1  |
>   2 |   2    2    2    2  |
>   3 |   3    3    3    3  |
>   4 |   4    4    4    4  |
>   5 |   5    5    5    5  |
>   6 |   6    6    6    6  |
>   7 |   7    7    7    7  |
>   8 |   8    8    8    8  |
>   9 |   9    9    9    9  |
>  10 |  10   10   10   10  |
>  11 |  11   11   11   11  |
>  12 |  12   12   12   12  |
>  13 |  13   13   13   13  |
>  14 |  14   14   14   14  |
>  15 |  15   15   15   15  |
>  16 |  16   16   16   16  |
>  17 |  17   17   17   17  |
>  18 |  18   18   18   18  |
>  19 |  19   19   19   19  |
>  20 |  20   20   20   20  |
>     +---------------------+
>
> which can be generalised for "any value of 20" (or 4 for that matter).
> On the other hand, this behaviour of J() was not in the very first
> release of Mata
>
> SJ-8-3  pr0043  . . . . . Stata tip 67: J() now has greater replicating powers
>        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
>        Q3/08   SJ 8(3):450--451                                 (no commands)
>        tip detailing Mata J() function recent enhancements


That is certainly worth reading, but in my case there is no functional
dependency
between the element's index (in the matrix) and its value. I just have
792 numbers
(and in general case more) to store in a (792 by 1) matrix. You can
think of them
as model coefficients estimates that are predefined and need to be
stored securely
in Mata code.

The above 1,1,1,1\2,2,2,2... line was just an example of the notation
(is there an
official name for this notation of spelling out a matrix's content?).

I find
http://www.stata.com/help.cgi?limits
one of the most important Stata documentation parts. But I can't find its Mata
counterpart. Perhaps, some restrictions are mentioned in the documentation here
and there, but not in such a crisp and clear form.

Thank you, Sergiy



>
> and so may not be old-fashioned enough for Sergiy. But there are other
> ways to do it. Even the pedestrian
>
> (1..20)',(1..20)',(1..20)',(1..20)'
>
> looks a better solution to me. What am I missing?
>
> Nick
>
> On Tue, Jan 18, 2011 at 3:44 PM, Sergiy Radyakin <[email protected]> wrote:
>> Hello Kit,
>>
>> On Tue, Jan 18, 2011 at 9:23 AM, Christopher Baum <[email protected]> wrote:
>>> <>
>>> On Jan 18, 2011, at 2:33 AM, Sergiy wrote:
>>>
>>>> What is the best way to define, e.g. 100000x1 matrix in the Mata code?
>>>> (Another software is writing the Mata program, so I am not concerned
>>>> about the amount of labor).
>>>
>>> Why not just define the matrix (or the vec() of the matrix) as a Stata variable and view it in Mata?
>>
>>
>> Because it goes against the encapsulation paradigm ("a language
>> mechanism for restricting access
>> to some of the object's components" -- Wikipedia).
>>
>> My two other observations are:
>> 1) if I try to input a matrix with 1,1,1,1\2,2,2,2\3,3,3,3.... it hits
>> the limit as well with a different error
>> message, but still pretty soon to be practical.
>>
>> 2) if I run the original commands without wrapping them into a
>> function - everything works well
>> (presumably because the literals are not stored, but rather executed
>> right away).
>>
>>
>> I have found by now the response of David Drukker from 2007 regarding
>> the problem:
>> http://www.stata.com/statalist/archive/2007-08/msg00149.html
>> Since I need the program to be compatible with Stata 9, further
>> updates will not be applicable.
>>
>> Thank you, Sergiy
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index