search

JingxuanH / xtune

R package 📦 for feature-specific tuning integrating external information based on elastic-net model
Other
0 stars 2 forks source link

Error: requires at least two non-NA values for interpolation #1

Open Tianmaru opened 1 month ago

Tianmaru commented 1 month ago

Calling xtune will sometimes succeed, at other times result in an error:

fit <- xtune(X = X_train, Y = y_train, Z = priors, family = "linear")
> Error in approx(lambda, seq(lambda), sfrac) : requires at least two non-NA values for interpolation

R version 4.2.2 xtune version: 2.0.0

Tianmaru commented 3 weeks ago

I got another error:

Start estimating alpha:
Error in optim(alpha.old, likelihood.alpha.theta.xtune, likelihood.alpha.theta.gradient.xtune,  : 
  non-finite value supplied by optim