Augmented Dickey Fuller

Overview


Augmented Dickey Fuller



let residuals = transformed.map(p=>{
  return p.y - ($obj(p).dot(reg.Coefficients) + reg.Coefficients.intercept)
})
Try it!
(see Hill, Griffiths, Lim pg 486)
Model 1% 5% 10%
{% \Delta y_t = \gamma y_{t-1} + v_t %} -2.56 -1.94 -1.62
{% \Delta y_t = \alpha + \gamma y_{t-1} + v_t %} -3.43 -2.86 -2.57
{% \Delta y_t = \alpha +\lambda t + \gamma y_{t-1} + v_t %} -3.96 -3.41 -3.13
Standard normal -2.33 -1.65 -1.28

Contents