It uses 'tidyeval' and 'dplyr' to create a linear
regression model.

`linear_regression_db(df, y_var = NULL, sample_size = NULL, auto_count = FALSE)`

## Arguments

- df
A Local or remote data frame

- y_var
Dependent variable

- sample_size
Prevents a table count. It is only used for models
with three or more independent variables

- auto_count
Serves as a safeguard in case sample_size is not
passed inadvertently. Defaults to FALSE. If it is ok for the
function to count how many records are in the sample, then set to
TRUE. It is only used for models with three or more independent variables

## Details

The linear_regression_db() function only calls one of three unexported functions.
The function used is determined by the number of independent variables. This is
so any model of one or two variables can use a simpler formula, which in turn
will have less SQL overhead.

## Examples

```
library(dplyr)
mtcars %>%
select(mpg, wt, qsec) %>%
linear_regression_db(mpg)
#> # A tibble: 1 × 3
#> `(Intercept)` wt qsec
#> <dbl> <dbl> <dbl>
#> 1 19.7 -5.05 0.929
```