“Error in fitter…” While trying to use lrm function from rms library

Question

I'm trying to run a small data table through the lrm function from the rms library, but every time I run it I get this:

Error in fitter(X, Y, offset = offs, penalty.matrix = penalty.matrix,  : 
NA/NaN/Inf in foreign function call (arg 1)

Here's some reproducible code:

library(tidyverse)
library(rms)

data <- tribble(
~a, ~b, ~c,
"name", 2, 77.95535,
"name", 2, 81.36152,
"name", 2, 80.87081,
"name", 2, 80.59054,
"name", 2, 79.36807,
"name", 2, 82.42083,
"name", 2, 78.80646,
"name", 2, 78.88949
)

lmod <- lrm(b~c, data, penalty=0, x=T, y=T, tol=1e-10, maxit=1e6)

I've tried a variety of things. I tried switching b~c to c~b (I know it wouldn't yield the same data) and I got an error, but not the same error. I tried using only the first two parameters, as well as a variety of including or excluding the following ones. I've also tried debugging and following the code into the lrm function, and I narrowed it down to (shocker) the fitter function.

Here's some resources that I've been using that you might find helpful for yourselves or to help me. If there's anything I forgot to include drop a comment and I'll fill in what I can:) Thanks!

R Reference Manual

lrm Function Description

R for Data Science (book)

lrm Function Source Code

Answer 1

I get the same error using your supplied data. Then I changed some b to 1, the error disappeared:

data <- tribble(
  ~a, ~b, ~c,
  "name", 2, 77.95535,
  "name", 2, 81.36152,
  "name", 1, 80.87081,
  "name", 1, 80.59054,
  "name", 2, 79.36807,
  "name", 1, 82.42083,
  "name", 1, 78.80646,
  "name", 2, 78.88949
)

lmod <- lrm(b~c, data = data, penalty=0, x=TRUE, y=TRUE, tol=1e-14, maxit=1e10)
lmod

Logistic Regression Model

 lrm(formula = b ~ c, data = data, x = TRUE, y = TRUE, penalty = 0, 
     tol = 1e-14, maxit = 1e+10)

                   Model Likelihood    Discrimination    Rank Discrim.    
                         Ratio Test           Indexes          Indexes    
 Obs         8    LR chi2      0.00    R2       0.000    C       0.500    
  1          4    d.f.            1    g        0.000    Dxy     0.000    
  2          4    Pr(> chi2) 1.0000    gr       1.000    gamma   0.000    
 max |deriv| 3                         gp       0.000    tau-a   0.000    
                                       Brier    0.250                     

           Coef   S.E.    Wald Z Pr(>|Z|)
 Intercept 0.0000 39.9600 0.00   1.0000  
 c         0.0000  0.4992 0.00   1.0000 

I know it's only a subset of your data, but maybe all rows in b have value 2? You can check that with eg, table(data$b)