A new method, ChiMIC, to calculate the MIC values
Source code of the ChiMIC
2) Make sure that c++ has installed in your computer for compilation; In the matlab runtime environment, use the mex -setup command to set: <use of'Microsoft Visual C++' for C++ language compilation>
3) Running program “make.m” to compile the equipartitionYaxis2.c, getsuper2var.c and getmutualI2var_fix4.c to mex files;
The first column of data is Y (dependent variable), the rest of the columns (X) independent variable;
make;
num=randperm(size(data,1));
data=data(num',:);% scramble the samples
sample_num=size(data,1);
[MIC,bestc,mycertain,certain_seg,mutual_I_1,mutual_I_2]=MIC_OIC_chi_1_1(data,sample_num^0.55,5,sample_num);
[MIC,~]=MIC_OIC_chi_1_1_class(data,0.55,5)
Source: Chen Yuan, Zeng Ying, Luo Feng*, Yuan Zheming*. A New Algorithm to Optimize Maximal Information Coefficient. PLoS ONE, 2016 11(6): e0157567