time-series-analysis（时间序列分析）

Time Series Analysis

When to use

• Building a forecast for operational planning (staffing, inventory, infrastructure capacity)
• Identifying whether a trend is genuine or driven by seasonality
• Detecting anomalies in a metric stream (traffic spikes, revenue dips, error rate surges)
• Providing a "what would have happened" baseline for measuring initiative impact
• Presenting year-over-year growth in a way that accounts for seasonal patterns

Process

1. Load and inspect the time series — confirm regular intervals (fill gaps if needed), check for obvious data quality issues (negative values, zeros in non-zero series), and identify the natural granularity (daily, weekly, monthly).
2. Test for stationarity — run an ADF test. If non-stationary (trend or seasonality present), note this — it informs decomposition and model choice rather than blocking analysis. See references/ts_patterns_guide.md.
3. Decompose into components — separate the time series into trend, seasonal, and residual using additive or multiplicative decomposition. Measure the strength of each component (0–1). Strong seasonality (>0.6) means raw values are misleading without seasonal adjustment.
4. Detect anomalies — flag points more than 3 standard deviations from the rolling median. Investigate the top 5 anomalies against the event log (product releases, campaigns, incidents). Use scripts/ts_analyzer.py --detect-anomalies.
5. Fit a forecast model — fit an ARIMA model (or simpler moving average if data is short). Validate on a held-out 20% test set and report MAPE. Generate point estimates and 95% confidence intervals for the forecast horizon.
6. Produce the analysis report — summarise trend direction and strength, seasonal patterns and their business implications, anomaly findings, and the forecast with uncertainty. Use assets/ts_report_template.md.

Inputs the skill needs

• Time series data: date column + one numeric metric column, minimum 2 full seasonal cycles
• Granularity of the data (daily, weekly, monthly)
• Forecast horizon required (days, weeks, months ahead)
• Event log or change log for anomaly investigation
• Business context: what drives this metric, known seasonal patterns

Output

• scripts/ts_analyzer.py — decomposes, detects anomalies, and fits an ARIMA forecast; outputs charts and CSV
• references/ts_patterns_guide.md — stationarity, seasonality types, model selection guide, and common pitfalls
• assets/ts_report_template.md — report template: characteristics, decomposition summary, anomaly list, forecast table, insights