sparsenormalise.m.svn-base

来自「a function inside machine learning」· SVN-BASE 代码 · 共 18 行

function [normalisedX] = sparseNormalise(X)
% Normalize the features (columns) of sparse matrix X

[numRows, numCols] = size(X); 
normalisedX = sparse(numRows, numCols); 
normCols = speye(numCols); 

for i=1:numCols
    column = X(:, i);
    normCol = norm(column);
    
    if normCol ~= 0 
        normCols(i, i) = 1/normCol; 
    end 
end

normalisedX = X*normCols;