How do I create a large binary target matrix?

2 次查看(过去 30 天)
Hello. I have a dataset of 300 samples (with 60 observations each) that are divided into 6 even categories (50 samples each)
For neural networking purposes, I would like to create a target matrix such that:
0-50 = category 1
51-50 = category 2
.
.
.
251-300 = category 6
I plan to assign these values by creating a 6x300 matrix. and placing a 1 in row 1 for 0-50, a 1 in row 2 for 51-100 and so on. I believe I may be able to do this by manipulating the eye function, but I have not seen examples of this. I know there must be an easier way to do this than by logging the numbers manually... any ideas? thanks

采纳的回答

Image Analyst
Image Analyst 2017-6-15
Seems kind of weird, but going precisely by your directions, 6 lines of code should do it - one line of code for the 6 groups. Can't get much simpler or easier than that. See lines of code below, following your directions which are comments:
% creating a 6x300 matrix.
m = rand(6, 300);
% placing a 1 in row 1 for 0-50,
m(1, 1:50) = 1;
% a 1 in row 2 for 51-100,
m(2, 51:100) = 1;
% and so on
m(3, 101:150) = 1;
m(4, 151:200) = 1;
m(5, 201:250) = 1;
m(6, 251:300) = 1;
The only difference is I started at column 1 instead of 0 because there is no column 0.

更多回答(2 个)

Walter Roberson
Walter Roberson 2017-6-14
编辑:Walter Roberson 2017-6-14

Greg Heath
Greg Heath 2017-6-15
The only way to use eye is
target = repmat(eye(6),50)
Otherwise start with
target = zeros(6,300)
then insert ones(1,50) into the correct locations.
Hope this helps.
Thank you for formally accepting my answer
Greg

Community Treasure Hunt

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

Start Hunting!

Translated by