Refactor for better sklearn compatibility

This is a major API refactor, intended to accomplish the following:

- Reduce the number of named objects and concepts, simplifying the
  user-facing API.
- Improve compatibility with `sklearn`. The goal here is not to be a
  drop-in API for `sklearn`, instead the goal is to be able to integrate
  `ibisml` into `sklearn` workflows. With this refactor you should be
  able to include a `ibisml.Recipe` as a step within an existing
  `sklearn` pipeline and have things "just work".
- Still support usage outside of sklearn workflows. We don't want to
  force users to fit the sklearn model, but given sklearn is a common
  standard within the python ecosystem, it'd be good to make the common
  case simpler.

.pre-commit-config.yaml

@@ -1,13 +1,8 @@
 repos:
-  - repo: https://github.com/psf/black
-    rev: 22.10.0
-    hooks:
-      - id: black
-        language_version: python3
-
   - repo: https://github.com/astral-sh/ruff-pre-commit
     # Ruff version.
-    rev: v0.0.270
+    rev: v0.3.1
     hooks:
       - id: ruff
-        args: [--fix, --exit-non-zero-on-fix]
+        args: [--fix]
+      - id: ruff-format

ibisml/__init__.py

@@ -1,5 +1,5 @@
 from ._version import __version__
-from ibisml.core import Recipe, RecipeTransform, Step, Transform, TransformResult
+from ibisml.core import Recipe, Step
 from ibisml.select import (
     selector,
     everything,
@@ -22,4 +22,3 @@
     where,
 )
 from ibisml.steps import *  # noqa: F403
-from ibisml import transforms