function X = chisquare_inv(P,V);
%CHISQUARE_INV Inverse of chi-square cumulative distribution function (cdf).
%
% X = chisquare_inv(P,V) returns the inverse of chi-square cdf with V
% degrees of freedom at fraction P.
% This means that P*100 percent of the distribution lies between 0 and X.
%
% To check, the answer should satisfy: P==gammainc(X/2,V/2)
% Uses FMIN and CHISQUARE_SOLVE
%
% Written January 1998 by C. Torrence
if (nargin < 2), error('Must input both P and V');, end
if ((1-P) < 1E-4), error('P must be < 0.9999');, end
if ((P==0.95) & (V==2)) % this is a no-brainer
X = 5.9915;
return
end
MINN = 0.01; % hopefully this is small enough
MAXX = 1; % actually starts at 10 (see while loop below)
X = 1;
TOLERANCE = 1E-4; % this should be accurate enough
vers = version;
vers = str2num(vers(1));
while ((X+TOLERANCE) >= MAXX) % should only need to loop thru once
MAXX = MAXX*10.;
% this calculates value for X, NORMALIZED by V
% Note: We need two different versions, depending upon the version of Matlab.
if (vers >= 6)
X = fminbnd('chisquare_solve',MINN,MAXX,optimset('TolX',TOLERANCE),P,V);
else
X = fmin('chisquare_solve',MINN,MAXX,[0,TOLERANCE],P,V);
end
MINN = MAXX;
end
X = X*V; % put back in the goofy V factor
return
% end of code