DoseResponse / drc

Fitting dose-response models in R
https://doseresponse.github.io/drc/
21 stars 16 forks source link

summary(m) does not work with error: Error in solve.default(object$fit$hessian) : Lapack routine dgesv: system is exactly singular: U[1,1] = 0 #36

Open lijuanyanbasf opened 9 months ago

lijuanyanbasf commented 9 months ago

dataframe <- data.frame(conc=c(0, 0.944, 2.18, 4.14, 8.37, 16.1), total=c(160, 80, 80, 80, 80, 80), response=c(3, 1, 1, 5, 8, 80))

1) summary(m) does not work as below (increasing trend)

m <- drm(response/total ~ conc, weights=total, data=dataframe, fct = W1.4(), type = "binomial") plot(m) summary(m)

**# Error in solve.default(object$fit$hessian) :

Lapack routine dgesv: system is exactly singular: U[1,1] = 0**

2) However, summary(m) works if I model the non-responding fractions (decreasing trend)

m <- drm((total-response)/total ~ conc, weights=total, data=dataframe, fct = W1.4(), type = "binomial") plot(m) summary(m)