Add support for mgcv::gam()

[juliasilge]

Closes #131

library(vetiver)
library(mgcv)
#> Loading required package: nlme
#> This is mgcv 1.8-40. For overview type 'help("mgcv-package")'.
training_sim <- gamSim(1, n = 50, scale = 2)
#> Gu & Wahba 4 term additive model

model_fit <- mgcv::gam(y ~ s(x1, bs = 'ps', sp = 0.6) + 
                           s(x2, bs = 'ps', sp = 0.6) + x3, 
                       data = training_sim, method="REML")
summary(model_fit)
#> 
#> Family: gaussian 
#> Link function: identity 
#> 
#> Formula:
#> y ~ s(x1, bs = "ps", sp = 0.6) + s(x2, bs = "ps", sp = 0.6) + 
#>     x3
#> 
#> Parametric coefficients:
#>             Estimate Std. Error t value Pr(>|t|)    
#> (Intercept)   8.4262     0.7017  12.009  3.9e-14 ***
#> x3           -0.9844     1.2596  -0.782     0.44    
#> ---
#> Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> 
#> Approximate significance of smooth terms:
#>         edf Ref.df     F  p-value    
#> s(x1) 6.056  6.813 6.377 8.48e-05 ***
#> s(x2) 6.028  6.802 7.155 2.58e-05 ***
#> ---
#> Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> 
#> R-sq.(adj) =  0.689   Deviance explained = 77.2%
#> -REML = 117.71  Scale est. = 4.4707    n = 50

v <- vetiver_model(model_fit, "simulated-gam")
v
#> 
#> ── simulated-gam ─ <butchered_gam> model for deployment 
#> A generalized additive model (gaussian family, identity link) using 3 features

library(plumber)
pr() %>% vetiver_api(v)
#> # Plumber router with 2 endpoints, 4 filters, and 1 sub-router.
#> # Use `pr_run()` on this object to start the API.
#> ├──[queryString]
#> ├──[body]
#> ├──[cookieParser]
#> ├──[sharedSecret]
#> ├──/logo
#> │  │ # Plumber static router serving from directory: /Library/Frameworks/R.framework/Versions/4.2-arm64/Resources/library/vetiver
#> ├──/ping (GET)
#> └──/predict (POST)

^{Created on 2022-08-17 with reprex v2.0.2}

We need to fix up the butcher method for mgcv::gam(), though.