Vectorizing for loop for faster processing

1 view (last 30 days)
Ifeoluwa Ajiboye
Ifeoluwa Ajiboye on 10 Mar 2022
Edited: AndresVar on 10 Mar 2022
I have a column of sorted row numbers labelled "Row Numbers"
Row Numbers =
1
3
5
8
100
...
...
I want to use information from Row numbers to sort a table labbeled "Table" into different sections
Table =
Time SensorOut
1:00 1
2:00 3
3:00 2.3
4:00 3.4
5:00 3.1
6:00 30
7:00 3
8:00 3
9:00 3
... ...
... ...
... ...
What i want to achieve is this
Table =
Time SensorOut Section
1:00 1 1
2:00 3 1
3:00 2.3 2
4:00 3.4 2
5:00 3.1 3
6:00 30 3
7:00 3 3
8:00 3 3
9:00 3 4
... ... ...
... ... ...
... ... ...
What i have done so far
for i = 1:length(rownumbers)-1
Table.Section(rownumbers(i):rownumbers(i+1)) = i;
end
This works but the issue is that i have 100,000 rows of data and each loop is taking about 0.4 seconds. I was wondering if any one had an idea how to vectorize this for loop for faster processing? Or a more clever way of achiveing this?

Accepted Answer

AndresVar
AndresVar on 10 Mar 2022
Edited: AndresVar on 10 Mar 2022
You defined bin edges so you can try discretize: Group data into bins or categories - MATLAB discretize (mathworks.com)
idx=1:10
idx = 1×10
1 2 3 4 5 6 7 8 9 10
rowNums=[1 3 5 8 inf]
rowNums = 1×5
1 3 5 8 Inf
discretize(idx,rowNums)'
ans = 10×1
1 1 2 2 3 3 3 4 4 4

More Answers (0)

Community Treasure Hunt

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

Start Hunting!