以下的這個程序是用來做方差比檢驗的m-file 大家看看有問題么

我跑出來的結果只有一個數，而我要得到兩個數的，即同方差和異方差下的方差比

function [VR,Zk,Zhk]=vrt_full(x,k)

%

% Syntax: [vr,zk,zhk]=vrt(x,k)

%

% Calculates the Variance Ratio Test (VR Test) of a time series x, with

% and without the heteroskedasticity assumption.

% input??: x is a vector of time serie (observed prices)

%? ?? ?? ? k is a scalar

% output : VR is the value of the VRTest

%? ?? ?? ? Zk is the homoscedastic statistic of the variance ratio

%? ?? ?? ? Zhk is the heteroscedastic statistic of the variance ratio

%

x=xlsread('D:\A股\數\book1.xls');

k=2;

rt1=diff(x); % one period rate of return

T=length(rt1);

i=1:T-k+1;

M=zeros(T-k+1,k);

for j=1:k,

M(:,j)=i+j-1;

end

rtk=sum(rt1(M'));??% k period rate of return

moy=mean(rt1);

v=var(rt1);

m=k*(T-k+1)*(1-k/T);

VR=1/m*sum((rtk-k*moy).^2)/v; % Variance ratio statistic

Zk=sqrt(T)*(VR-1)*(2*(2*k-1)*(k-1)/(3*k))^(-.5); % homoscedastic statistic

j=1:k-1;

vec1=(2/k*(k-j)).^2;

rst=(rt1-moy).^2;

aux=zeros(1,k-1);

for i=1:k-1,

aux(i)=rst(i+1:T)'*rst(1:T-i);

end

vec2=aux/((T-1)*v)^2;

Zhk=(VR-1)*(vec1*vec2')^(-.5); % heteroscedastic statistic

總結

以上是生活随笔為你收集整理的matlab做数据间方差,用matlab做方差比检验的问题的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。