swiping row in Matrix

i have this matrix (M)

M=[
      1 NaN NaN 0.9 0.3; 
      2 14.1450 5.142 0.85 -0.1145; 
      3 NaN NaN NaN NaN; 
      4 NaN NaN 1.25 0.5;    
      5 18.45 9.3507 1.63 0.049; 
      6 NaN NaN NaN NaN;
      7 17.16 0 0.7163 0.2791;  
      8 NaN NaN 1 0.35;
      9 NaN NaN NaN NaN]

i want to reshape like this

P=[
      1 1 17.16 0 0.7163 0.2791; 
      2 6 18.45 9.3507 1.63 0.049; 
      3 3 14.1450 5.142 0.85 -0.1145; 
      4 5 NaN NaN 1.25 0.5; 
      5 2 NaN NaN 0.9 0.3; 
      6 8 NaN NaN 1 0.35;
      7 4 NaN NaN NaN NaN; 
      8 7 NaN NaN NaN NaN; 
      9 9 NaN NaN NaN NaN ]

NOTE: in first matrix 3rd coloum 7 row it has zero that why it came 1st

earlier it was 9x5 then it convert into 9x6

how can i make this?

[Walter Roberson]

You seem to be doing a sort of some kind, but I do not see what sorting rule you are using? You mention column 3 but your second row has a higher value for that column than the third row does? And I do not understand where that second column is coming from ??

[ANAND ASP]

in output 1st column indicate just an serial number

2nd column indicate previous data's serial number

Rule: in 3rd column "0" must be first others are randomly

then comes with all are filled up rows

then comes with filled data in 4th and 5th colunm

then NaN

But row must be same do not change the data of row, but just swap it up and down

[ANAND ASP]

increasing order is doesnt matter, but the data should not change, just swape as it is whole row,

you can set as in increasing order

[Walter Roberson]

2nd column indicate previous data's serial number

That does not appear to be the case. The previous serial number for the 18.45 row was 5, but it becomes 6 in the output. The previous serial number for the 17.16 line was 7, but it becomes 1 in the output.

[ANAND ASP]

P=[ 1 7 17.16 0 0.7163 0.2791;

2 2 14.1450 5.142 0.85 -0.1145;

3 5 18.45 9.3507 1.63 0.049;

4 1 NaN NaN 0.9 0.3;

5 8 NaN NaN 1 0.35;

6 4 NaN NaN 1.25 0.5;

7 3 NaN NaN NaN NaN;

8 6 NaN NaN NaN NaN;

9 9 NaN NaN NaN NaN ]

now this output data with no confusion

[ANAND ASP]

M= [1.0000 NaN NaN 0.9000 0.3000;

2.0000 14.1450 NaN 0.8500 -0.1145;

3.0000 NaN NaN NaN NaN;

4.0000 NaN NaN 1.2500 0.5000;

5.0000 18.4500 9.3507 1.6300 0.0490;

6.0000 NaN NaN NaN NaN;

7.0000 17.1600 0 0.7163 0.2791;

8.0000 13.1500 NaN 1.0000 NaN;

9.0000 NaN NaN NaN NaN]

[~,is]=sortrows([sum(isnan(M),2) any(M==0,2)],[1 -2])

P=M(is,:)

P=[(1:size(P,1))', P]

using this equations, i get this,

P =

1.0000 7.0000 17.1600 0 0.7163 0.2791

2.0000 5.0000 18.4500 9.3507 1.6300 0.0490

3.0000 2.0000 14.1450 NaN 0.8500 -0.1145

4.0000 1.0000 NaN NaN 0.9000 0.3000

5.0000 4.0000 NaN NaN 1.2500 0.5000

6.0000 8.0000 13.1500 NaN 1.0000 NaN

7.0000 3.0000 NaN NaN NaN NaN

8.0000 6.0000 NaN NaN NaN NaN

9.0000 9.0000 NaN NaN NaN NaN

but i expect this

P =

1.0000 7.0000 17.1600 0 0.7163 0.2791

2.0000 5.0000 18.4500 9.3507 1.6300 0.0490

3.0000 2.0000 14.1450 NaN 0.8500 -0.1145

4.0000 8.0000 13.1500 NaN 1.0000 NaN

5.0000 1.0000 NaN NaN 0.9000 0.3000

6.0000 4.0000 NaN NaN 1.2500 0.5000

7.0000 3.0000 NaN NaN NaN NaN

8.0000 6.0000 NaN NaN NaN NaN

9.0000 9.0000 NaN NaN NaN NaN

{ 1st priority is 4th column: data must be in sequence (0,data,...,data,NaN,....NaN) }

{ 2rnd priority is 3rd column: data must be in sequence (data,...,data,NaN,...,NaN) }

{ 3rd priority is 5th & 6th column: data must be in sequency(data,.....data,NaN,...NaN)}

{ 4th priority for all "NaN" terms in others row }

{ Data of row should not be change just swap the row with priority}

{ one more value is needed, in above matrix 3rd column hsve 4 data so i want 4 as output (A=4)}

Answer 1

[M.Many]

Try this :

M= [1.0000       NaN       NaN    0.9000    0.3000;
    2.0000   14.1450       NaN    0.8500   -0.1145;
    3.0000       NaN       NaN       NaN       NaN;
    4.0000       NaN       NaN    1.2500    0.5000;
    5.0000   18.4500    9.3507    1.6300    0.0490;
    6.0000       NaN       NaN       NaN       NaN;
    7.0000   17.1600         0    0.7163    0.2791;
    8.0000   13.1500       NaN    1.0000       NaN;
    9.0000       NaN       NaN       NaN       NaN]

[P,index] = sortrows(M,3)

P(isnan(P(:,3)),:) = sortrows(P(isnan(P(:,3)),:),2)
P(isnan(P(:,2)),:) = sortrows(P(isnan(P(:,2)),:),4)
P(isnan(P(:,4)),:) = sortrows(P(isnan(P(:,4)),:),5)

P = [ [1:9]' P]