Dr Cyril Pernet - 2012-08-01

Robust correlation toolbox functions

Data visualization

  1. corr_normplot(X,Y); % scatter plot and histrograms
  2. density = joint_density(X,Y); % joint density

Assumption checking

  1. HZmvntest([X Y]); % test normality
    % function from Trujillo-Ortiz et al - Matlab exchange
  2. [h,CI]\ = variance_homogeneity(X,Y); % test homoscedasticity
  3. outliers = detect_outliers(X,Y); % returns univariate and bivariate outliers

Correlations

  1. [r,t,p,hboot,CI]\ = Pearson(X,Y); % Pearson
  2. [r,t,p,hboot,CI]\ = Spearman(X,Y); % Spearman
  3. [r,t,p,hboot,CI]\ = bendcorr(X,Y); % percentage bend correlation
  4. [r,t,h,outid,hboot,CI]\ = skipped_correlation(X,Y); % skipped correlation

See the manual for details about options - or type in matlab help \<name_of_the_function></name_of_the_function>