[英]Speed-efficient classification for complex vectors in MATLAB
I am trying to optimize this piece of code and get rid of the nested loop implemented. 我正在尝试优化这段代码并摆脱实现的嵌套循环。 I am finding difficulties in applying a matrix to pdist function 我发现将矩阵应用于pdist函数时遇到了困难
For example, 1+j // -1+j // -1+j // -1-j are the initial points and i am trying to detect 0.5+0.7j to with point it belong by min distance approach . 例如,1 + j // -1 + j // -1 + j // -1-j是初始点,我试图通过最小距离逼近来检测0.5 + 0.7j。
any help is appreciated 任何帮助表示赞赏
function result = minDisDetector( newPoints, InitialPoints)
result = [];
for i=1:length(newPoints)
minDistance = Inf;
for j=1:length(InitialPoints)
X = [real(newPoints(i)) imag(newPoints(i));real(InitialPoints(j)) imag(InitialPoints(j))];
d = pdist(X,'euclidean');
if d < minDistance
minDistance = d;
index = j;
end
end
result = [result; InitialPoints(index)];
end
end
You can use efficient euclidean distance calculation as listed in Speed-efficient classification in Matlab
for a vectorized solution
- 您可以使用Speed-efficient classification in Matlab
中的Speed-efficient classification in Matlab
列出的高效欧氏距离计算,用于vectorized solution
-
%// Setup the input vectors of real and imaginary into Mx2 & Nx2 arrays
A = [real(InitialPoints) imag(InitialPoints)];
Bt = [real(newPoints).' ; imag(newPoints).'];
%// Calculate squared euclidean distances. This is one of the vectorized
%// variations of performing efficient euclidean distance calculation using
%// matrix multiplication linked earlier in this post.
dists = [A.^2 ones(size(A)) -2*A ]*[ones(size(Bt)) ; Bt.^2 ; Bt];
%// Find min index for each Bt & extract corresponding elements from InitialPoints
[~,min_idx] = min(dists,[],1);
result_vectorized = InitialPoints(min_idx);
Quick runtime tests with newPoints
as 400 x 1
& InitialPoints
as 1000 x 1
: 使用newPoints
作为400 x 1
和InitialPoints
为1000 x 1
快速运行时测试:
-------------------- With Original Approach
Elapsed time is 1.299187 seconds.
-------------------- With Proposed Approach
Elapsed time is 0.000263 seconds.
The solution is very simple. 解决方案非常简单。 However you do need my cartprod.m function to generate a cartesian product. 但是,您确实需要我的cartprod.m函数来生成笛卡尔积。
First generate random complex data for each variable. 首先为每个变量生成随机复杂数据。
newPoints = exp(i * pi * rand(4,1));
InitialPoints = exp(i * pi * rand(100,1));
Generate the cartesian product of newPoints
and InitialPoints
using cartprod
. 使用cartprod
生成newPoints
和InitialPoints
的笛卡尔积。
C = cartprod(newPoints,InitialPoints);
The difference of column 1 and column 2 is the distance in complex numbers. 第1列和第2列的差异是复数的距离。 Then abs
will find the magnitude of the distance. 然后abs
会找到距离的大小。
A = abs( C(:,1) - C(:,2) );
Since the cartesian product is generated so that it permutates newPoints
variables first like this: 由于生成笛卡尔积,因此它首先置换newPoints
变量,如下所示:
1 1
2 1
3 1
4 1
1 2
2 2
...
We need to reshape
it and get the minimum using min
to find the minimum distance. 我们需要reshape
它并使用min
来获得最小值以找到最小距离。 We need the transpose to find the min for each newPoints
. 我们需要转置来为每个newPoints
找到min。 Otherwise without the transpose, we will get the min for each InitialPoints
. 否则没有转置,我们将获得每个InitialPoints
。
[m,i] = min( reshape( D, length(newPoints) , [] )' );
m
gives you the min, while i
gives you the indices. m
给你min,而i
给你指数。 If you need to get the minimum initialPoints
, just use: 如果您需要获得最小的initialPoints
,只需使用:
result = initialPoints( mod(b-1,length(initialPoints) + 1 );
It is possible to eliminate the nested loop by introducing element-wise operations using the euclidean norm for calculating the distance as shown below. 通过使用欧几里德范数引入逐元素运算来计算距离,可以消除嵌套循环,如下所示。
result = zeros(1,length(newPoints)); % initialize result vector
for i=1:length(newPoints)
dist = abs(newPoints(i)-InitialPoints); %calculate distances
[value, index] = min(dist);
result(i) = InitialPoints(index);
end
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.