Improving preprocessing

[aPovidlo]

This is a 🔨 code refactoring.

Summary

Significant Updates in Data Storage and Preprocessing

Major Updates:

• Enhanced logging: Added more detailed logs in DEBUG mode during preprocessing.
• New functionality: You can now mark categorical features in data when using InputData.from_numpy(...), InputData.from_dataframe(...), and InputData.from_csv(...) methods.
• New class: Introduced OptimizedFeatures, which stores data with optimal dtypes for improved efficiency.
• Preprocessing improvement: Added a new stage called reduce_memory_size to optimize memory usage.
• API enhancements: Updated PredefinedModel to allow copying parameters from DataPreprocessor.

Minor Updates:

• Improved logic for detecting categorical data.
• Updated encoders and imputers to align with the new changes.
• Revised tests to incorporate the new features.

Context

closes #1337 closes #1329