create scaffold polars prototype

[isabelizimm]

Notes:

import polars as pl
from vetiver import VetiverModel, VetiverAPI
from vetiver.data import mtcars
from sklearn.linear_model import LinearRegression

schema = {
    "mpg": pl.Float64,
    "cyl": pl.Int32,
    "disp": pl.Float32,
    "hp": int,
    "drat": float,
    "wt": float,
    "qsec": float,
    "vs": bool,
    "am": bool,
    "gear": pl.Int32,
    "carb": int,
}

mtcars = pl.DataFrame(mtcars, schema=schema)

model = LinearRegression().fit(mtcars.drop(columns="mpg"), mtcars["mpg"])
v = VetiverModel(model, model_name = "cars_linear", 
                 prototype_data = mtcars.drop(columns="mpg"))

VetiverAPI(v).run()

• right now, this will not coerce or check polars types, rather, it will use Python types (eg, if you try to pass "hello" as cyl: you get ValueError: could not convert string to float: 'hello')
• need to refactor to not have polars as a dependency
• add tests

[isabelizimm]

closing, potentially looking at pandera at a later point to support this instead