K-Dense-AI/scientific-agent-skills / skills/timesfm-forecasting
timesfm-forecasting
VerifiedPerforms zero-shot time-series forecasting with Google's TimesFM, including regular-grid CSV preparation, quantile forecasts, XReg covariates, and held-out evaluation. Uses the Apache-licensed TimesFM 2.5 checkpoint by default and documents the distinct TimesFM 3.0 multivariate API and weight-license requirements.
examples/global-temperature/README.md
View on GitHub# Illustrative temperature-shaped forecast
The small bundled CSV has no recorded retrieval/revision provenance. It is
retained as illustrative numerical input, not as verified NASA GISTEMP or NOAA
observations. No claim about 2025 climate, a cooling trend or an anomalous
September is justified from this example. Use a cited versioned authoritative
data release for scientific analysis.
The scripts now use the Apache-licensed 2.5 checkpoint with package 3.0.2 and
correct deciles: mean at index 0, q10..q90 at 1..9, nominal 80%/60% bands.
Old generated CSV/JSON/PNG/GIF/HTML outputs were removed. New output requires
explicit regeneration; no pretrained inference was executed in this refresh.
From the skill root, after installing dependencies and checking resources:
```bash
python examples/global-temperature/run_forecast.py --output-dir /tmp/timesfm-demo
python examples/global-temperature/visualize_forecast.py --output-dir /tmp/timesfm-demo
python examples/global-temperature/generate_animation_data.py --output-dir /tmp/timesfm-demo
python examples/global-temperature/generate_gif.py --output-dir /tmp/timesfm-demo
python examples/global-temperature/generate_html.py --output-dir /tmp/timesfm-demo
```
Animations forecast from each prefix only. All-history/final-forecast background
layers are labeled retrospective references and are never model inputs. Band
axes use all steps, avoiding clipping wider early forecasts. New-schema checks
reject old unvalidated animation/forecast JSON rather than silently relabel it.
The HTML embeds data but loads pinned Chart.js from a CDN, so it needs network
access and is not a self-contained offline artifact. Its interpolation is linear
to avoid suggesting extra forecast structure between monthly observations.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| deprecation-and-migrationaddyosmani/agent-skillsManages deprecation and migration. Use when removing old systems, APIs, or features. Use when migrating users from one implementation to another. Use when migrating a database schema in production, such as renaming or dropping a column without downtime (expand/contract). Use when deciding whether to maintain or sunset existing code. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| performance-optimizationaddyosmani/agent-skillsOptimizes application performance across frontend, backend, queries, and databases. Use when performance requirements exist, when you suspect performance regressions, when Core Web Vitals or load times need improvement, when N+1 query patterns need fixing, or when profiling reveals bottlenecks. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| agent-cost-reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json. | Stars: 118k | Updated | Claude CodeCodex | Unverified | Install |
| agent-cost-reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json. | Stars: 118k | Updated | Claude CodeCodex | Unverified | Install |