Chi-Square Test

Chi-Square Test for continuous distributions.

Description

[A, B] = CHI2TEST(DATA, N, ALPHA, DIST, X, Y, Z) returns the chi-square statistic for the samples contained in the row vector DATA.

N specifies the number of equal-probability class intervals for the test. ALPHA is the confidence level parameter used to find the critical chi-square value.

DIST is a string containing the probability distribution that we are testing against. See the staitsctics toolbox for supported distributions - 'exp', 'gam', 'unif' are some of them.

X, Y, and Z specify the estimated parameters for the selected DIST. Some distributions require only one of these parameters, and the order that these parameters are provided follows the values given to the cummulative distribution functions UNIFCDF, GAMCDF, EXPCDF, and others.

A is the computed chi-square statistic, and B is the critical tabulated value at the degrees of freedom. The degree of freedom is the number of intervals minus the number of estimated parameters.
 
In general, if A is less than B, the H0 hypothesis that DATA follows the DIST distribution is accepted.

An attempt to fit some data with the uniform distribution on the interval from 1.5 to 2.9. The test fails, since A > B:

[a, b] = chi2test (data, 10, 0.05, 'unif', 1.5, 2.9)
   a =
      38.7500
   b =
      14.0671
   
See also MLE, CHI2INV, CHI2STAT, HIST, CDF, ICDF, PDF

• Required Products: Statistics Toolbox
• MATLAB release: MATLAB 6.1 (R12.1)

[fishwine huang]

thank you

[elsa m]

I get this error:
??? One or more output arguments not assigned during call to 'chi2test'.

what am i doing wrong??

[Tshiamo Motshegwa]

>>[a, b] = chi22test (out_data, 70, 0.05, 'unif',28580,5076)
??? Error using ==> minus
Matrix dimensions must agree.

Error in ==> chi22test at 100
chi2bins = ((o_freq - e_freq).^2)./e_freq;

[Reuben Stahls]

Hi,

Was wondering if you could reply to this question. I have a Gaussian random vector with a non-zero mean and I would like to compute the probability that the frobenius of this random vector is less than some constant. Is there any available matlab function that can help me achieve my goal.

Thanks for your response in advance.

Best wishes,
Reuben Stahls

[Nasir Ali]

good

[manj benning]

0_freq is a row vec and e_freq is a column vec...so i had to modify line 100
chi2bins = ((o_freq' - e_freq).^2)./e_freq;
to make it work.

[JianfeiWu Wu]

 

[Rimvydas]

I think the line for calculation of p value should be corrected
from
pval=1-gammainc(x2/2, (length(O)-3)/2); %p value
to
pval=1-gammainc(x2/2, (length(O)-1)/2); %p value
since degrees of freedom as gamma function parameter was incorrectly defined

[Zuhal]

i have wind speed data including 365 values. how can we define the "n" value?