How to do Principle Component Analysis (PCA)

Finding literature on Principle Component Analysis (PCA) is very easy but very hard to find exactly how to do or implement PCA. Hence, I have demonstrated simple MATLB code below. Following code assumes that variable in columns and numbers of records in row. Variables name are long because I want those to be self descriptive.

covResult = cov(trainingInputs); % Calculating covariance matrix
[V,D] = eig(covResult); % Calculating eigen values and eigen vector
PCATrainingInputs = trainingInputs * V(:,HowManyMostVariantPrincipleComopnenetWantToUse:NumberOfInputVariables);