deming linear estimation
res = deming1(y,x,lambda,alpha)
* y = (N x 1) vector of the endogenous variable measured with error
* x = (N x 1) vector of the endogenous variable measured with error
* lambda = a scalar, the assumed ratio of the measurement errors of y and x: sigma2_y / sigma2_x
* alpha = a scalar, the confidence level
* res = a results tlist with
- res('meth') = 'deming'
- res('y') = the y vector of endogenous variables
- res('x') = the x vector of exogenous variables
- res('errors variance ratio') = the assumed ratio of the measurement errors of y and x
- res('Estimated x values') = the estimated "true" x values
- res('Estimated y values') = the estimated "true" y values
- res('regression st. error') = Standard error of regression estimate
- res('coeff st. error') = estimate of standard error of the slope and intercept
- res('tstat') = Sutnedt's T of the estimated coefficients
- res(string(100*alpha)+'% confidence interval') = the confidence interval of the coefficients at the chosen level
// provides the deming regression presented at: // http://www.real-statistics.com/regression/deming-regression/deming-regression-basic-concepts/ load(GROCERDIR+'\deming_d.dat') res = deming1(y,x,0.4,0.95) | ![]() | ![]() |