Scalecast

Added
- Added `init_dates` and `levely` attributes
- Added `'Observations'` info to history and `export()`
- Added `'lvl_test_set_predictions'` to export dataframes
Changed
- Got rid of `first_obs` and `first_dates` attributes and wrote more efficient code to do what they were there for
- More information available when `__str__()` is called
Fixed
- Fixed what became an issue with the last update in which when calling `add_diffed_terms()` or `add_lagged_terms()`, the level series wasn't accurate due to how undifferencing was being executed. After examining this issue, it became evident that the previous way to undifference forecasts was less efficient than it should have been, this update fixed the issues from the last update and made the code more efficient
- Fixed an issue where AR terms were manipulating the underlying xreg structures so that each forecast were using its own test-set propogated AR values instead of the correct AR values
- Fixed the `export_Xvars_df()` method which wasn't working correctly if at least one forecast hadn't been called first

Added
- added the following functions that can each add additional Xvars to forecast with:
- `add_exp_terms()` - for non polynomial exponential transformations
- `add_logged_terms()` - for log of any base transformations
- `add_pt_terms()` - for individual variable power transformations (box cox and yeo johnson available)
- `add_diffed_terms()` - to difference non-y terms
- `add_lagged_terms()` - to lag non-y terms
- added the 'pt' normalizer for yeo-johnson normalization (in addition to 'minmax', 'normalize', and 'scale')
- added the `drop_Xvars()` function that is identical to the `drop_regressors()` function
Changed
- imports all sklearn models as soon as scalecast is imported
- src code cleanup with better coding practices when it comes to forecasting sklearn models (no more copying and pasting new functions)
- changed several set data types to lists in src code
- changed the names of some hidden functions
- other src code cleanup for readability and minor efficiency gains
- better in-line comments and docstring documentation
- got rid of quiet paramater in `save_summary_stats()` and `save_feature_importance()` and now these simply log any problems as warnings
- time trends now start at 1 instead of 0 (makes log transformations possible)
- observation dropping for AR terms in sklearn models now based on the number of N/A values in each AR term instead of just the AR number
- changed some example grids to include the pt normalizer
Fixed
- now logs all warnings

Added
- `dynamic_testing` argument to `manual_forecast()` and `auto_forecast()` functions -- this is `True` by default (makes all testing comparable between sklearn/non-sklearn models)
- `dynamic_tuning` argument to `tune()` function -- this is `False` by default to majorly improve speed in some applications
Changed
- native Forecaster warnings will be logged
Fixed

Added
- added `tune_test_forecast()` function to notebook module to create a progress bar when using a notebook
Changed
Fixed
- fixed an issue with `Forecaster.ingest_Xvars_df()` when `use_future_dates=False` causing an error to be raised

Added
- added `include_traing` parameter to `notebook.results_vis()` function
Changed
Fixed
- fixed `print_attr` parameter default in `notebook.results_vis()`

Added
- added `results_vis()` notebook function (requires ipywidgets)
- added `Forecaster.export_Xvars_df()` function
- added `max_integration` argument to the `Forecaster.integrate()` function
Changed
Fixed