How may I calculate the number of rows in each group?

2 Ansichten (letzte 30 Tage)
Julia
Julia am 13 Feb. 2023
Kommentiert: Julia am 13 Feb. 2023
Hi! If I have a matrix:
M= [1 50 60 70 50 40
2 NaN 10 20 10 10
3 NaN 20 NaN NaN NaN
1 NaN 60 30 40 50
2 10 20 10 20 NaN
1 30 20 40 NaN 50
2 NaN 50 50 NaN NaN]
The first column indicates to which group the rows belong to. "1, 2, 3" (row 1, 2, 3) are Group 1, "1, 2" (row 4, 5 ) are Group 2, the next "1, 2" ( row 6, 7) are Group 3.
I am trying to find the number of rows in each group. The desired result would be:
N= [3
2
2]
Could someone please teach me how to do this? It seems that I can't get my head around it. I think the first step can be splitting the groups:
g=nan(size(M,1),1);
ix=(M(:,1)==1);
g(ix)=[1:sum(ix)];
g=fillmissing(g,'previous');
but I am not sure where to go from here.
  2 Kommentare
KSSV
KSSV am 13 Feb. 2023
Whats the logic for this?
"1, 2, 3" (row 1, 2, 3) are Group 1, "1, 2" (row 4, 5 ) are Group 2, the next "1, 2" ( row 6, 7) are Group 3
Dyuman Joshi
Dyuman Joshi am 13 Feb. 2023
Bearbeitet: Dyuman Joshi am 13 Feb. 2023
So when the value in first column is 1, a new group starts? If so, @Voss's answer should be good to go.
If not, then specify a general criteria for splitting into groups.

Melden Sie sich an, um zu kommentieren.

Akzeptierte Antwort

Voss
Voss am 13 Feb. 2023
diff(find([M(:,1); 1] == 1))

Weitere Antworten (1)

Bhavana Ravirala
Bhavana Ravirala am 13 Feb. 2023
Hi Julia,
According to my understanding, the first element of the row is deciding the group. You can use the below code to know the number of rows in each group.
N=M(:,1); % getting the first column of elements which we are using to divide the groups
k=length(N); % gives the number of rows in the matrix
j=[];
o=0;
for i=1:k
if(n(i)==1)
if(o~=0)
j=[j,o];
end
o=1;
else
o=o+1;
if i==k
j=[j,o];
end
end
end
Hope this helps!!

Tags

Produkte


Version

R2022b

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by