Time Series Aggregation and Forecasting Starter
Aggregate events and produce a baseline forecast.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
Aggregate event data and build a baseline forecast I can trust. Data description (columns, time range, granularity): [DATA_DESCRIPTION]. Language and libraries: [LANGUAGE_AND_LIBRARIES]. What I need to forecast and the horizon: [FORECAST_GOAL]. Seasonality I expect: [SEASONALITY]. Provide a plan as tables first: 1. Steps: Step | Purpose | Method | Check. 2. Data issues to check: Issue | Test | Handling (missing periods, duplicates, outliers, time zones, daylight saving, late-arriving data). 3. Aggregation choices: Level (hourly, daily, weekly, monthly) | Pros and cons | Recommended for my goal. 4. Baseline models: Method (last value, moving average, seasonal naive, exponential smoothing) | When it works | Expected weakness. 5. Evaluation: train and test split by time (never random), error measures (mean absolute error, percentage error with the zero problem noted), and a backtest with rolling origin. Then provide the code: load, clean, resample, plot, build the seasonal naive baseline, a simple exponential smoothing model, evaluate, and output the forecast with an interval. Explain that a model must beat the seasonal naive baseline to be worth using. Add a short guide to reading the result for non-technical stakeholders and a list of limits (small data, structural changes).
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.