It's not likely you'll find one library that contains all functions you'd like to use in your code as there are tons of different machine learning algorithms, most with their own dedicated package.
There's no problem in using functions from different packages in your code, as you will see that's often the case. If you don't want to load a whole library, you can always install the package and then only load the function of interest by using the package's name, then type
:: then the function.
myMLE = stats4::mle(minuslogl, start = formals(minuslogl), method = "BFGS",
fixed = list(), nobs, …)
You can also use the https://www.rdocumentation.org/ webstite to search for packages/functions and see how popular they are compared to others (they use percentile for this)
Hope this helps,