Any questions from a PROFI to a SUPER PROFI - 1. - page 26

New comment
 
FOReignEXchange:


if if and if again. I would start with a1a1

Yes, I can do that, but we need to analyze not the equality, but some similarity or say, how much the data in the matrices are "by eye" similar and not even how much the data itself, but their coordinates in the matrix
 
FOReignEXchange:

so there's no other way out but logic.
there is a way out, i already googled the topic and neural networks and went into the maze of cryptography - but i hope to find an easier way to find matrix correlation i guess
 
IgorM:
yes, I can do that, but it's not the equality, but a certain similarity or let's say how much the data in the matrices are "by eye" similar and not even how much the data itself, but their coordinates in the matrix

try subtracting one mariage element by element from the other.

you'll get something like.

| 0,0,0,0,0,0, 0,0,0,0,0,0 |
| 0,0,1,0,0,0, 0,0,0,0,0,0 |
| 0,0,0,0,0,0, 0,0,0,0,-1,0 |

and then you can play around with this simplified

 
sergeev: try subtracting one mariage element by element from the other.

Oh! That's great advice!

Thanks! I'll try to think along those lines.

 
IgorM:
There is a way out, I already googled the topic and neural networks and went into the maze of cryptography - but I hope to find an easier way to search for matrix correlation probably


I've forbidden myself to go to this thread, moreover - to reply, it looks like a VIP-club...

On the subject: Borland's Paradox worked quite well for about twenty years :)

Well, if you want to do it yourself (I'm very fond of it) - look for optimal collapsing, - like a signature. Should help.

 

subtract one matrix from another and then analyse the non-zero cells...

Too late...

"Degree of non-similarity" = number of cells not equal to zero * modulus of sum of values

 
xrust:"Degree of dissimilarity" = number of cells not equal to zero * modulus of sum of values
oops, for a quick (rough) assessment will do
 
IgorM:

I can't even formulate a query for a search engine :(((

I want to use some method to compare like? correlation? several matrices with numerical data.

like this:

matrix1: matrix2:

| 4,4,1,2,1,1, 1,1,1,1,1,1 | | 4,4,1,2,1,1, 1,1,1,1,1,1 |
| 4,4,1,2,1,1, 1,1,1,1,1,1 | | 4,4,1,1,1,1, 1,1,1,1,1,1 |
| 4,4,4,2,1,1, 1,1,1,1,1,1 | | 4,4,4,2,1,1, 1,2,1,1,1,1 |

the given matrices have minimal differences, how can this be analysed programmatically?


The simplest way is the Minimum Sum of Squares of Differences. The closer to zero, the more accurate
 
As probably the only person here who got a C for matrices, I've cleaned up the flooding.
I remind you that moderation on the forum is present, but in this thread it is done in the form of strict thematic censorship.
 

Question about memory mapping.

Is it possible, without resorting to copying and recreating, to dynamically change the size of allocated memory (CreateFileMapping) and its projection (MapViewOfFile) ?

That is, the problem is as follows:

A CreateFileMapping object is created in memory to exchange data between processes (writer-reader) 100 bytes. and MapViewOfFile of the same size of 100 bytes.

The first process writing can write all 100 bytes of data into memory which the second process reading hasn't managed to save.

Therefore, is it possible to expand the allocated memory size without re-creating CreateFileMapping / MapViewOfFile again?
So that the first process would not wait for release and continued to write into the added space, while the second process continued to read further too.

1...192021222324252627282930313233...45
New comment