[R-sig-Geo] GWmodel package error

Metastate Metastate metastate at gmail.com
Sat Aug 29 16:37:56 CEST 2015


I am applying the GWmodel R package on my GWR analysis of the impact of
weather on the county-level crop yield. There are  about 2342 counties in
my analysis. But I got 2 errors for some variables as shown below. Would
you please help me out? Really appreciate for your any kindly help or

(1). Does the error below occur because the max_temperature_above 87F
(max_Tsfc_GT_86F_8) did not have enough unique (heterogeneous) levels? If
so, What is the minimum unique levels an independent variable should has to
avoid such error?

[1] "max_Tsfc_GT_86F_8"

[1] 30: the number of the unique levels of "max_Tsfc_GT_86F_8"

Error in solve.default(t(X * W.i) %*% X) :

  system is computationally singular: reciprocal condition number =

There is another similar one:

[1] "min_Tsfc_LT_32F_8"

[1] 6

Error in solve.default(t(X * W.i) %*% X) :

  Lapack routine dgesv: system is exactly singular: U[2,2] = 0

(2). One county in my dataset does not have value of "Qsoil_40_100_7". So,
I guess the error is due to the missing value. Am I right?

[1] "Qsoil_40_100_7"

[1] 1338

Error in cov.wt(matrix(y, ncol = 1), wt = rep(as.numeric(1), dp.n),
method = "ML") :

  length of 'wt' must equal the number of rows in 'x'

Really appreciate!


	[[alternative HTML version deleted]]

More information about the R-sig-Geo mailing list