I am trying to do a stepwise regression to find the best predictors for my model but when running the code in r I run into an error.

I used the step() function on my data set but I got the error

reg<-lm(depression~.,data=dataset)
stepAIC(reg, direction="both")

>Error in stepAIC(reg, direction = "both") : 
  AIC is -infinity for this model, so 'stepAIC' cannot proceed

I am using a data set of counties in the US and corresponding demographic and geograpic data. The dataset has 3135 observations so I am only including the first 6 rows.

> head(dataset)
# A tibble: 6 × 17
  county    state depression FIPS   size  total black  white asian latino native pacificIslander other mixed diversity density classification
  <chr>     <chr>      <dbl> <chr> <dbl>  <dbl> <dbl>  <dbl> <dbl>  <dbl>  <dbl>           <dbl> <dbl> <dbl>     <dbl>   <dbl> <chr>         
1 Abbeville SC          22.3 45001  491.  24368  6372  16658    58    441     28               0   138   673     0.463    49.6 Small         
2 Acadia    LA          27.3 22001  655.  57674  9511  44413    95   1780     13               0    40  1822     0.378    88.0 Small         
3 Accomack  VA          18.8 51001  449.  33367  9446  19813   268   3084     56               0    55   645     0.558    74.3 Small         
4 Ada       ID          20.3 16001 1052. 497494  5815 411348 12672  44317   1668             990  2384 18300     0.306   473.  Medium        
5 Adair     IA          16.6 19001  569.   7479    48   7046     9    196      0              12    49   119     0.111    13.1 Small         
6 Adair     KY          26.1 21001  405.  18887   307  17308    59    468      3              24     6   712     0.158    46.6 Small         

Does anyone know what is causing this and how I can fix it?

(For reproducibility)

> dput(head(dataset))
structure(list(county = c("Abbeville", "Acadia", "Accomack", 
"Ada", "Adair", "Adair"), state = c("SC", "LA", "VA", "ID", "IA", 
"KY"), depression = c(22.3, 27.3, 18.8, 20.3, 16.6, 26.1), FIPS = c("45001", 
"22001", "51001", "16001", "19001", "21001"), size = c(491.195, 
655.244, 449.32, 1052.015, 569.271, 405.292), total = c(24368, 
57674, 33367, 497494, 7479, 18887), black = c(6372, 9511, 9446, 
5815, 48, 307), white = c(16658, 44413, 19813, 411348, 7046, 
17308), asian = c(58, 95, 268, 12672, 9, 59), latino = c(441, 
1780, 3084, 44317, 196, 468), native = c(28, 13, 56, 1668, 0, 
3), pacificIslander = c(0, 0, 0, 990, 12, 24), other = c(138, 
40, 55, 2384, 49, 6), mixed = c(673, 1822, 645, 18300, 119, 712
), diversity = c(0.463214524775288, 0.377844145794569, 0.558287177093522, 
0.306246361818157, 0.111453827636757, 0.157904845203141), density = c(49.6096255051456, 
88.0191195951432, 74.2611056707914, 472.896299007143, 13.1378552569866, 
46.600969177778), classification = c("Small", "Small", "Small", 
"Medium", "Small", "Small")), row.names = c(NA, -6L), class = c("tbl_df", 
"tbl", "data.frame"))