Can I Pass an Object to Next Exercise Chunk?

[dudrnjs1391]

I executed this code. But, All chunk Separated and cant't inherited from previous chunk.
Can I use the data from the previous chunk to the next chunk?

• Make an Object and Check It

"```{r 2.1.data_create, exercise=TRUE}```"

"```{r 2.1.data_create-solution, exercise.setup="2.1.data_create-solution"}
tmp <- c(1,2,3)
```"

"```{r 2.2.data_plus, exercise=TRUE, ref.label="2.1.data_create-solution"}```"

"```{r 2.2.data_plus-solution}
tmp <- tmp+1
```"

the result will be tmp = 2,3,4

"```{r 2.3.data_plus_plus, exercise=TRUE, ref.label="2.2.data_plus-solution"}```"

"```{r 2.3.data_plus_plus-solution}
tmp <- tmp+2
tmp
```"

the result will be tmp = 4,5,6

[jjallaire]

That's correct. Each exercise chunk stands alone. This can definitely be an annoying limitation, however there are a few reasons why we do this:

1. So that users can run exercises in any order they wish;

2. So that exercises are "stateless" and can as a result run in a more scalable fashion (e.g. creating a farm of servers that just run exercies)

3. We automatically save the state of tutorials between runs (including outputs). It's cost prohibitive storage wise though to save the actual R objects (as they could be data frames of unbounded size) so the user just sees their outputs replayed. In this case the actual R state is lost and exercises depending on previous exercises would be broken.

The workaround is to create setup chunks that already have the "expected" values primed.