system is computationally singular: reciprocal condition number

1

I'm dealing with time series in R. My data are homocedastic, have no autocorrelation, have no unit root at lag = 0, trend or seasonality.

I still have a problem with this specific error:

    VAR_S <- summary(VAR(Base_Dados,p = 1, type = "both"))
  

Error in solve.default (Sigma):

     

system is computationally singular: reciprocal condition number =
          5.67068e-37

VAR<- (VAR(Base_Dados, p = 1, type = "both" ) )

At first I thought it was collinearity, but the correlation of the data does not exceed 0.25.

    
asked by anonymous 03.05.2018 / 13:31

1 answer

0

The matrix can not be invertible, the determinant of the matrix is zero or approximately zero. I handled this in two different ways, change the data to a similar model that worked badly and did not show the desired result for analysis. as for the previous model the resolution form was re-evaluated the scale of the variables they ceased to be in the same unit escalation but started to have approximately the same number of integers per observation.

It was an intuitive work, so I have no formulas to demonstrate.

    
07.05.2018 / 05:33