Should the LHS be checked for interaction terms? Other things?

[DavisVaughan]

No description provided.

[DavisVaughan]

library(hardhat)
library(gapminder)

# warning? same as rhs when indicators=FALSE?
mold(country:continent ~ year, gapminder)
#> $predictors
#> # A tibble: 1,704 x 1
#>     year
#>    <dbl>
#>  1  1952
#>  2  1957
#>  3  1962
#>  4  1967
#>  5  1972
#>  6  1977
#>  7  1982
#>  8  1987
#>  9  1992
#> 10  1997
#> # … with 1,694 more rows
#> 
#> $outcomes
#> # A tibble: 1,704 x 2
#>    country     continent
#>    <fct>       <fct>    
#>  1 Afghanistan Asia     
#>  2 Afghanistan Asia     
#>  3 Afghanistan Asia     
#>  4 Afghanistan Asia     
#>  5 Afghanistan Asia     
#>  6 Afghanistan Asia     
#>  7 Afghanistan Asia     
#>  8 Afghanistan Asia     
#>  9 Afghanistan Asia     
#> 10 Afghanistan Asia     
#> # … with 1,694 more rows
#> 
#> $preprocessor
#> Formula Preprocessor: 
#>  
#> # Predictors: 1 
#>   # Outcomes: 2 
#>    Intercept: FALSE 
#>   Indicators: TRUE 
#> 
#> $offset
#> NULL

mold(year ~ country:continent, gapminder, indicators = FALSE)
#> Warning: Interaction terms have been detected in `formula`. These are
#> not expanded when `indicators = FALSE`, but the individual terms will be
#> included in the output.
#> $predictors
#> # A tibble: 1,704 x 2
#>    country     continent
#>    <fct>       <fct>    
#>  1 Afghanistan Asia     
#>  2 Afghanistan Asia     
#>  3 Afghanistan Asia     
#>  4 Afghanistan Asia     
#>  5 Afghanistan Asia     
#>  6 Afghanistan Asia     
#>  7 Afghanistan Asia     
#>  8 Afghanistan Asia     
#>  9 Afghanistan Asia     
#> 10 Afghanistan Asia     
#> # … with 1,694 more rows
#> 
#> $outcomes
#> # A tibble: 1,704 x 1
#>     year
#>    <int>
#>  1  1952
#>  2  1957
#>  3  1962
#>  4  1967
#>  5  1972
#>  6  1977
#>  7  1982
#>  8  1987
#>  9  1992
#> 10  1997
#> # … with 1,694 more rows
#> 
#> $preprocessor
#> Formula Preprocessor: 
#>  
#> # Predictors: 2 
#>   # Outcomes: 1 
#>    Intercept: FALSE 
#>   Indicators: FALSE 
#> 
#> $offset
#> NULL

^{Created on 2019-02-16 by the reprex package (v0.2.1.9000)}

[DavisVaughan]

maybe check the attr(terms, "factor")` matrix for interactions?

[DavisVaughan]

Can use a modified version of detect_factors_in_interactions() to look for : in the factor matrix of the terms (* and ^ and %in% are all recoded with : in the factor matrix). This would let us prevent interactions on the LHS because only model.frame() is run.

[github-actions]

This issue 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.