calculate differences inside columms between absolute and dinamic references inside of Matrix. Saving results.

Dear friends, few days ago, I asked a question a litle bit similar than this one, but I made some changes and i need reply specially for this problem. In the other case I needed to re-arrange in advance the elements, now, I need to do the calculations shown below:

I have a matrix like this imput:

i mput:

    306  305,952  304,683  302,265  300,219 (1)
    306  306,008  304,977  302,45  299,846 (2)
    306  305,962  304,84  302,413  299,812 (3)
    306  306,051  305,051  302,396  299,877 (4)
    306  306,003  304,891  302,163  300,017 (5)
    306  306,004  305,002  301,997  300,186 (6)

    406  406,016  404,943  402,467  400,224 (7)
    406  406,009  404,961  402,483  400,361 (8)
    406  406,014  404,852  402,547  400,287 (9)
    406  406,043  404,99  402,52  400,372 (10)
    406  406,048  405,105  402,425  400,817 (11)
    406  406,034  404,928  402,425  400,623 (12)

    506  506,038  505,182  502,518  500,878 (13)
    506  506,003  505,208  502,728  500,684 (14)
    506  505,971  505,11  502,841  501,118 (15)
    506  505,979  505,083  502,644  500,922 (16)
    506  505,985  505,065  502,942  500,955 (17)
    506  506,062  505,067  502,818  501,28  (18)

I would like to calculate the differences between rows, this is, for example, all rows values minus the absoute one, but keeping order, I mean, create a new matrix with this differences. This, would be able with mxn matrix, not taking care about the size of this.

Such a example, in this link is the original file of imput

https://www.dropbox.com/s/y3ehqalm3cxmzuj/x.txt

    desired output --> out

    100  100,064  100,26  100,202  100,005 (7)-(1)
    100  100,001  99,984  100,033  100,515 (8)-(1)
    100  100,052  100,012  100,134  100,475 (9)-(1)
    100  99,992  99,939  100,124  100,495 (10)-(1)
    100  100,045  100,214  100,262  100,8   (11)-(1)
    100  100,03  99,926  100,428  100,437 (12)-(1)

    200  200,086  200,499  200,253  200,659 (13)-(1)
    200  199,995  200,231  200,278  200,838 (14)-(1)
    200  200,009  200,27  200,428  201,306 (15)-(1)
    200  199,928  200,032  200,248  201,045 (16)-(1)
    200  199,982  200,174  200,779  200,938 (17)-(1)
    200  200,058  200,065  200,821  201,094 (18)-(1)

Depending on the result I will obtain I will post a new question, related with this one but not directly connected with.

Best Regards , and heartfelt thanks.

Answer 1

[Azzi Abdelmalek]

 A=[306 305.952 304.683 302.265 300.219 
    306  306.008  304.977  302.45  299.846 
    306  305.962  304.84  302.413  299.812 
    306  306.051  305.051  302.396  299.877 
    306  306.003  304.891  302.163  300.017 
    306  306.004  305.002  301.997  300.186 
    406  406.016  404.943  402.467  400.224 
    406  406.009  404.961  402.483  400.361 
    406  406.014  404.852  402.547  400.287 
    406  406.043  404.99  402.52  400.372 
    406  406.048  405.105  402.425  400.817 
    406  406.034  404.928  402.425  400.623 
    506  506.038  505.182  502.518  500.878 
    506  506.003  505.208  502.728  500.684 
    506  505.971  505.11  502.841  501.118 
    506  505.979  505.083  502.644  500.922 
    506  505.985  505.065  502.942  500.955 
    506  506.062  505.067  502.818  501.28 ]

out=bsxfun(@minus,A(7:end,:),A(1,:))

[Miguelcm90]

If i increase the number of rows with the same number in the first column,

out = bsxfun(@minus,x(7:end,:),x(1,:))

should be changed, right?, I need something which detects when the number of the first column is changing.

Thanks