R: Object not found

0
votes

Sales is a variable in the dataset train. Summary() finds it, but not glm(). What's going on?

Any help is appreciated.

> summary(train$Sales)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
      0    3727    5744    5774    7856   41550 
> summary(ross_glm <- glm(Sales ~ Customers,family=Gamma,data = train[Sales>0])) 
Error in `[.data.frame`(train, Sales > 0) : object 'Sales' not found

P.S. I've tried referencing using train$Sales and also attaching the train dataset, but they don't resolve the problem.

1
r
Your subset is not correct. It should be something like train[train$Sales>0,]. See for example: set.seed(42); train <- data.frame(Sales = rnorm(100), Other = rnorm(100)); train[Sales>0]. - user3710546
you could also use summary(ross_glm <- glm(Sales ~ Customers,family=Gamma,data = train, subset = Sales>0) - Ben Bolker

1 Answers

4
votes

Edit: Ben Bolker's comment provides a cleaner way to subset inside the function glm.

There is problem in the way you subset your data.frame. It should be train[train$Sales>0,] (don't forget the , after the filter).

set.seed(42)
train <- data.frame(Sales = rnorm(100), Other = rnorm(100))
train[Sales>0]

Error in [.data.frame(train, Sales > 0) : object 'Sales' not found

head(train[train$Sales>0,])
#       Sales      Other
# 1 1.3709584  1.2009654
# 3 0.3631284 -1.0032086
# 4 0.6328626  1.8484819
# 5 0.4042683 -0.6667734
# 7 1.5115220 -0.4222559
# 9 2.0184237  0.1881930