yes, thanks!
On Sun, Mar 31, 2013 at 12:47 AM, Jason Grout
wrote:
> On 3/30/13 4:09 PM, Gary McConnell wrote:
>
>> So I understand something of what is going on, could you please tell me
>> why the tuple thing is necessary for the "used" set but the vectors are
>> necessary for the "bases": ie w
One other thought - if we do manage to get this search routine to work
"sequentially" (ie not necessarily to have to hold all things in memory at
the same time), then perhaps we can extend this philosophy right back to
the definition of "N_1" (my candidate vectors with norm 1 etc), because
once aga
Hi again
well, not that I understand *how*, but that code works *magnificently*! I
see (confusingly for me I am afraid :) ) that the concept of "reserved
words" does not have much currency in SAGE ... !
The only difficulty with using that code is that for the sets I am looking
at, the number of b
On 3/30/13 4:56 PM, Gary McConnell wrote:
OK so thinking about it, even though your code is beautifully compact
and elegant, I think I am going to have to revert a little to the
"outer-inner-loop" structure in order to achieve what I need. Namely,
instead of storing "used" vectors, I store "used"
On 3/30/13 4:09 PM, Gary McConnell wrote:
So I understand something of what is going on, could you please tell me
why the tuple thing is necessary for the "used" set but the vectors are
necessary for the "bases": ie why can we not use a vector structure for
"used"?
The bases is just a list of l
OK so thinking about it, even though your code is beautifully compact and
elegant, I think I am going to have to revert a little to the
"outer-inner-loop" structure in order to achieve what I need. Namely,
instead of storing "used" vectors, I store "used" bases and search through
the remaining orth
Thanks again Jason - that seems to do the trick - now I just have to figure
out the equivalent of the old BASIC "goto 10" to start your loop again!!
So I understand something of what is going on, could you please tell me why
the tuple thing is necessary for the "used" set but the vectors are
neces
On 3/30/13 1:43 PM, Gary McConnell wrote:
OK now I'm back in a familiar nightmare - if I set the vectors to be
immutable then I cannot treat them as vectors because they seem to be of
"NoneType" or something so have no length etc etc; however I cannot /not
/set them to be immutable if I wish to w
OK now I'm back in a familiar nightmare - if I set the vectors to be
immutable then I cannot treat them as vectors because they seem to be of
"NoneType" or something so have no length etc etc; however I cannot *not *set
them to be immutable if I wish to work through a list consisting of those
very
Brilliant - thanks Jason - sorry for delay
On Sat, Mar 30, 2013 at 2:04 PM, Jason Grout wrote:
> On 3/30/13 8:39 AM, Gary McConnell wrote:
>
>> One more thing Jason: the code you wrote, as far as I can see, does not
>> take account of "N", which I guess is your N=set(N1) of my "N1" - the
>> poin
On 3/30/13 8:39 AM, Gary McConnell wrote:
One more thing Jason: the code you wrote, as far as I can see, does not
take account of "N", which I guess is your N=set(N1) of my "N1" - the
point is I guess that if I pass N1 to the function instead of your F
then it should be ok (ie it only searches wi
One more thing Jason: the code you wrote, as far as I can see, does not
take account of "N", which I guess is your N=set(N1) of my "N1" - the point
is I guess that if I pass N1 to the function instead of your F then it
should be ok (ie it only searches within the subset N1 of the full vector
space)
fair enough - I'm not too clear myself on how to describe it sorry - but
for example, if you look at the code I posted in the .txt file, ideally the
vv_i,ww_i,xx_i,yy_i etc would be variables which would only be created if
the requisite antecedent vectors had been found (as effectively happens in
yes indeed that is a problem in the general case but in fact in my
experience with the particular things I am looking at is that either the
first such set works, or else none of them do so it may not be as much
of a problem in this case. Fingers crossed!
On Sat, Mar 30, 2013 at 1:09 PM, Jaso
On 3/30/13 7:58 AM, Gary McConnell wrote:
Am I to deduce from this that what I was originally trying to do (ie
telescoping recursive variables) is not possible in Python?
Sorry---it wasn't ever really clear to me how you were approaching the
problem and what you meant by "telescoping recursive
On 3/30/13 7:49 AM, Gary McConnell wrote:
in fact the construction of N1 precludes isotropic vectors - I should
have mentioned that. The xRy function effectively forces separate sets
of vectors to be mutually exclusive also. I have learnt a lot from your
example there - but unfortunately I cannot
Hi again - thanks for that excellent new function - it's amazingly compact
and I think it may well do what I need - I am about to try modifying it to
work in my context, and will let you know.
Am I to deduce from this that what I was originally trying to do (ie
telescoping recursive variables) is
Hi
in fact the construction of N1 precludes isotropic vectors - I should have
mentioned that. The xRy function effectively forces separate sets of
vectors to be mutually exclusive also. I have learnt a lot from your
example there - but unfortunately I cannot get to where I need to - I have
just rec
On 3/30/13 7:12 AM, Jason Grout wrote:
So to summarize your code, you're finding M=3 sets of d=3 mutually
orthogonal vectors so that each pair of sets satisfies some criteria
(your XRY function)? Do the sets have to be mutually exclusive? Your
code doesn't enforce this, but your xRy function may
On 3/30/13 6:34 AM, Gary McConnell wrote:
Hi Jason - thanks a lot - I have attached a text file which is just a
cleaned-up version of my code, with some explanations at the top for
functions whose details are not relevant to the question. Apologies -
you'll see it's a "neat mess", so to speak ...
Hi Jason - thanks a lot - I have attached a text file which is just a
cleaned-up version of my code, with some explanations at the top for
functions whose details are not relevant to the question. Apologies -
you'll see it's a "neat mess", so to speak
Kind regards
Gary
On Sat, Mar 30, 2013
On 3/30/13 5:08 AM, Gary McConnell wrote:
Hi Jason
fair question - I was trying to avoid including all the boring details
of my code! In the simplest case I have a search routine which looks for
M sets of d vectors inside a vector space of dimension N over a finite
field, whose dot products sati
Hi Jason
fair question - I was trying to avoid including all the boring details of
my code! In the simplest case I have a search routine which looks for M
sets of d vectors inside a vector space of dimension N over a finite field,
whose dot products satisfy a bunch of polynomial equations. My prob
On 3/29/13 6:38 PM, GaryMak wrote:
d = 5;
for ii in range(0,d):
for a[ii] in range(0,d):
print a[ii]^2 + ii;
What do you expect the output of this to be? What are you trying to
accomplish with this code?
Thanks,
Jason
--
You received this message because you are subscribed
I guess that's what I don't understand then: why can I not invoke a
structure like this? In fact I need to nest several levels of this type of
structure, so I would then define variables b_xx etc as b[a[ii]] and so on.
If that is not how these variable definitions work, could someone please
tell me
oops, didn't read carefully enough... that third line just doesn't
make much sense.
On Fri, Mar 29, 2013 at 4:49 PM, Tom Boothby wrote:
> Gary,
>
> The third line
>
> for a[ii] in range(0,d):
>
> should read
>
> for ii in range(0,d):
>
> On Fri, Mar 29, 2013 at 4:38 PM, GaryMak wrote:
>> Hi guys
Gary,
The third line
for a[ii] in range(0,d):
should read
for ii in range(0,d):
On Fri, Mar 29, 2013 at 4:38 PM, GaryMak wrote:
> Hi guys
>
> apologies as usual for how dumb this question's going to sound ... but how
> do I actually use the variables a[1] etc? What I was hoping to do was to u
Hi guys
apologies as usual for how dumb this question's going to sound ... but how
do I actually *use *the variables a[1] etc? What I was hoping to do was to
use them in the following way (once I had already apparently successfully
invoked Jason's routine), where d is a variable I would like to
On Jul 4, 8:05 pm, Jason Grout wrote:
> This question has come up often enough that I thought posting a short
> snippet here might be useful. The question is: how do I automatically
> generate variables based on indices (e.g., a[0], a[1], etc. being
> variables). Here is one way:
>
> class Var
29 matches
Mail list logo
