A data scientist is working on a time-series forecasting project within Databricks, encountering datasets with irregular time intervals and missing time points. Which Databricks feature or library would best address these challenges?