Axe terms environment in method for recipe steps

[juliasilge]

Closes #192

This PR changes the axe_env.step method to nuke the terms environment. I deleted all the methods that did only that, because they can use this method now.

I made all the dplyr-ish steps that had the same thing happening in their axe methods inherit straight from axe_env.step_arrange so we can change once in the future if needed. I also moved these around a bit in the file so these are together. I think these were in alphabetical order before so I can move them back in you prefer that, vs. similar steps together.

Here are results from the current main branch:

## similar to reported issue from OP
library(recipes)
#> Loading required package: dplyr
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
#> 
#> Attaching package: 'recipes'
#> The following object is masked from 'package:stats':
#> 
#>     step
library(butcher)

prepped_rec <- 
  recipe(Species ~ ., data = iris) %>% 
  step_normalize(all_numeric()) %>% 
  step_pca(all_numeric()) %>%
  prep()

prepped_rec <- butcher(prepped_rec)

prepped_plus <- function() {
  some_junk_in_the_environment <- runif(1e6) 
  
  prepped_rec <- 
    recipe(Species ~ ., data = iris) %>% 
    step_normalize(all_numeric()) %>% 
    step_pca(all_numeric()) %>% 
    prep()
  
  prepped_rec <- butcher(prepped_rec)
  prepped_rec
}

lobstr::obj_size(prepped_rec)
#> 32,224 B
lobstr::obj_size(prepped_plus())
#> 8,032,552 B

^{Created on 2021-05-14 by the reprex package (v2.0.0)}

And to compare, here are the results from this PR:

## similar to reported issue from OP
library(recipes)
#> Loading required package: dplyr
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
#> 
#> Attaching package: 'recipes'
#> The following object is masked from 'package:stats':
#> 
#>     step
library(butcher)

prepped_rec <- 
  recipe(Species ~ ., data = iris) %>% 
  step_normalize(all_numeric()) %>% 
  step_pca(all_numeric()) %>%
  prep()

prepped_rec <- butcher(prepped_rec)

prepped_plus <- function() {
  some_junk_in_the_environment <- runif(1e6) 
  
  prepped_rec <- 
    recipe(Species ~ ., data = iris) %>% 
    step_normalize(all_numeric()) %>% 
    step_pca(all_numeric()) %>% 
    prep()
  
  prepped_rec <- butcher(prepped_rec)
  prepped_rec
}

lobstr::obj_size(prepped_rec)
#> 24,632 B
lobstr::obj_size(prepped_plus())
#> 24,632 B

^{Created on 2021-05-14 by the reprex package (v2.0.0)}

[DavisVaughan]

Can you please add a NEWS bullet?

[DavisVaughan]

Thanks!

[github-actions]

This pull request has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.