# Timesfm Forecasting

> Pronóstico zero-shot de series temporales con el modelo fundacional TimesFM de Google, para cualquier serie univariada sin entrenar un modelo propio. Admite CSV/DataFrame/arrays, con pronósticos puntuales e intervalos de predicción.

Fuente: https://skillsagentes.com/skills/k-dense-ai/scientific-agent-skills/timesfm-forecasting
Markdown: https://skillsagentes.com/skills/k-dense-ai/scientific-agent-skills/timesfm-forecasting.md

Repositorio: https://github.com/K-Dense-AI/scientific-agent-skills
Autor: K-Dense-AI
Licencia: Apache-2.0 license
Actualizado: hace 28 días
Coste de contexto: 86 tok instalada, 3.9k tok al activarse, 525.2k tok con todos los archivos del bundle
Bundle: 31 archivos, 2.0 MB
Permisos que pide: read write edit bash

## Instalación

Un skill son archivos markdown: los mismos archivos valen para cualquier agente y lo único que cambia es el directorio de destino, es decir la bandera `--agent`. Añade `-g` para instalarlo en todos los proyectos de la máquina.

```bash
# Claude Code
npx -y skills add K-Dense-AI/scientific-agent-skills --skill timesfm-forecasting --agent claude-code
# Cursor
npx -y skills add K-Dense-AI/scientific-agent-skills --skill timesfm-forecasting --agent cursor
# Codex
npx -y skills add K-Dense-AI/scientific-agent-skills --skill timesfm-forecasting --agent codex
# Gemini CLI
npx -y skills add K-Dense-AI/scientific-agent-skills --skill timesfm-forecasting --agent gemini
# Windsurf
npx -y skills add K-Dense-AI/scientific-agent-skills --skill timesfm-forecasting --agent windsurf
# Cline
npx -y skills add K-Dense-AI/scientific-agent-skills --skill timesfm-forecasting --agent cline
```

## Qué hace

- Genera pronósticos zero-shot de series temporales univariadas sin entrenar un modelo propio, usando TimesFM 2.5
- Devuelve pronósticos puntuales y bandas de cuantiles (intervalos de predicción calibrados)
- Ejecuta una verificación previa obligatoria de RAM, GPU y disco antes de cargar el modelo
- Admite pronóstico por lotes de cientos o miles de series y forecasting con variables exógenas (XReg)
- Detecta anomalías usando los intervalos de cuantiles como bandas de predicción

## Cuándo usarla

- Pronosticar cualquier serie temporal univariada (ventas, sensores, energía, signos vitales, clima) sin entrenar un modelo
- Necesitas pronósticos probabilísticos con intervalos de predicción calibrados
- Tienes que pronosticar cientos o miles de series de forma eficiente en lote
- Prefieres un modelo fundacional en vez de ajustar manualmente parámetros de ARIMA/ETS

## Cuándo no

- Necesitas modelos estadísticos clásicos con interpretación de coeficientes (usa statsmodels)
- Necesitas clasificación o clustering de series temporales (usa aeon)
- Los datos son tabulares, no temporales (usa scikit-learn)

## Qué la activa

- "Pronostica esta serie de ventas mensuales con TimesFM"
- "Dame el intervalo de predicción al 90% para los próximos 24 periodos"
- "Detecta anomalías en esta serie usando los cuantiles de TimesFM"
- "Corre el chequeo del sistema antes de cargar TimesFM"

## Antes de instalar

- Ejecuta `python scripts/check_system.py` antes de cargar el modelo; requiere ~4 GB de RAM para TimesFM 2.5, y los pesos (~800 MB) se descargan de HuggingFace en el primer uso.
- Necesita en el PATH: python
- Variables de entorno: HF_HOME
- makes network requests
- reads environment config

## Archivos

- SKILL.md — 15 KB
- examples/anomaly-detection/detect_anomalies.py — 17 KB
- examples/anomaly-detection/output/anomaly_detection.json — 9 KB
- examples/anomaly-detection/output/anomaly_detection.png — 212 KB
- examples/covariates-forecasting/demo_covariates.py — 19 KB
- examples/covariates-forecasting/output/covariates_data.png — 448 KB
- examples/covariates-forecasting/output/covariates_metadata.json — 2 KB
- examples/covariates-forecasting/output/sales_with_covariates.csv — 7 KB
- examples/global-temperature/README.md — 6 KB
- examples/global-temperature/generate_animation_data.py — 5 KB
- examples/global-temperature/generate_gif.py — 7 KB
- examples/global-temperature/generate_html.py — 21 KB
- examples/global-temperature/output/animation_data.json — 130 KB
- examples/global-temperature/output/forecast_animation.gif — 776 KB
- examples/global-temperature/output/forecast_output.csv — 1 KB
- examples/global-temperature/output/forecast_output.json — 4 KB
- examples/global-temperature/output/forecast_visualization.png — 153 KB
- examples/global-temperature/output/interactive_forecast.html — 149 KB
- examples/global-temperature/run_example.sh — 1 KB
- examples/global-temperature/run_forecast.py — 5 KB
- examples/global-temperature/temperature_anomaly.csv — 591 B
- examples/global-temperature/visualize_forecast.py — 3 KB
- references/api_reference.md — 8 KB
- references/data_preparation.md — 7 KB
- references/examples_and_validation.md — 6 KB
- references/output_and_config.md — 4 KB
- references/performance_tuning.md — 2 KB
- references/system_requirements.md — 6 KB
- references/workflows.md — 4 KB
- scripts/check_system.py — 16 KB
- scripts/forecast_csv.py — 8 KB

## SKILL.md

Reproducido tal cual desde K-Dense-AI/scientific-agent-skills bajo Apache-2.0 license. Esta sección es el documento original y está en inglés.

# TimesFM Forecasting

## Overview

TimesFM (Time Series Foundation Model) is a pretrained decoder-only foundation model
developed by Google Research for time-series forecasting. It works **zero-shot** — feed it
any univariate time series and it returns point forecasts with calibrated quantile
prediction intervals, no training required.

This skill wraps TimesFM for safe, agent-friendly local inference. It includes a
**mandatory preflight system checker** that verifies RAM, GPU memory, and disk space
before the model is ever loaded so the agent never crashes a user's machine.

> **Key numbers**: TimesFM 2.5 uses 200M parameters (~800 MB on disk, ~1.5 GB in RAM on
> CPU, ~1 GB VRAM on GPU). The archived v1/v2 500M-parameter model needs ~32 GB RAM.
> Always run the system checker first.

## When to Use This Skill

Use this skill when:

- Forecasting **any univariate time series** (sales, demand, sensor, vitals, price, weather)
- You need **zero-shot forecasting** without training a custom model
- You want **probabilistic forecasts** with calibrated prediction intervals (quantiles)
- You have time series of **any length** (the model handles 1–16,384 context points)
- You need to **batch-forecast** hundreds or thousands of series efficiently
- You want a **foundation model** approach instead of hand-tuning ARIMA/ETS parameters

Do **not** use this skill when:

- You need classical statistical models with coefficient interpretation → use `statsmodels`
- You need time series classification or clustering → use `aeon`
- You need multivariate vector autoregression or Granger causality → use `statsmodels`
- Your data is tabular (not temporal) → use `scikit-learn`

> **Note on Anomaly Detection**: TimesFM does not have built-in anomaly detection, but you can
> use the **quantile forecasts as prediction intervals** — values outside the 90% CI (q10–q90)
> are statistically unusual. See the `examples/anomaly-detection/` directory for a full example.

## ⚠️ Mandatory Preflight: System Requirements Check

**CRITICAL — ALWAYS run the system checker before loading the model for the first time.**

```bash
python scripts/check_system.py
```

This script checks:

1. **Available RAM** — warns if below 4 GB, blocks if below 2 GB
2. **GPU availability** — detects CUDA/MPS devices and VRAM
3. **Disk space** — verifies room for the ~800 MB model download
4. **Python version** — requires 3.10+
5. **Existing installation** — checks if `timesfm` and `torch` are installed

> **Note:** Model weights are **NOT stored in this repository**. TimesFM weights (~800 MB)
> download on-demand from HuggingFace on first use and cache in `~/.cache/huggingface/`.
> The preflight checker ensures sufficient resources before any download begins.

```mermaid
flowchart TD
    accTitle: Preflight System Check
    accDescr: Decision flowchart showing the system requirement checks that must pass before loading TimesFM.

    start["🚀 Run check_system.py"] --> ram{"RAM ≥ 4 GB?"}
    ram -->|"Yes"| gpu{"GPU available?"}
    ram -->|"No (2-4 GB)"| warn_ram["⚠️ Warning: tight RAM<br/>CPU-only, small batches"]
    ram -->|"No (< 2 GB)"| block["🛑 BLOCKED<br/>Insufficient memory"]
    warn_ram --> disk
    gpu -->|"CUDA / MPS"| vram{"VRAM ≥ 2 GB?"}
    gpu -->|"CPU only"| cpu_ok["✅ CPU mode<br/>Slower but works"]
    vram -->|"Yes"| gpu_ok["✅ GPU mode<br/>Fast inference"]
    vram -->|"No"| cpu_ok
    gpu_ok --> disk{"Disk ≥ 2 GB free?"}
    cpu_ok --> disk
    disk -->|"Yes"| ready["✅ READY<br/>Safe to load model"]
    disk -->|"No"| block_disk["🛑 BLOCKED<br/>Need space for weights"]

    classDef ok fill:#dcfce7,stroke:#16a34a,stroke-width:2px,color:#14532d
    classDef warn fill:#fef9c3,stroke:#ca8a04,stroke-width:2px,color:#713f12
    classDef block fill:#fee2e2,stroke:#dc2626,stroke-width:2px,color:#7f1d1d
    classDef neutral fill:#f3f4f6,stroke:#6b7280,stroke-width:2px,color:#1f2937

    class ready,gpu_ok,cpu_ok ok
    class warn_ram warn
    class block,block_disk block
    class start,ram,gpu,vram,disk neutral
```

### Hardware Requirements by Model Version

| Model | Parameters | RAM (CPU) | VRAM (GPU) | Disk | Context |
| ----- | ---------- | --------- | ---------- | ---- | ------- |
| **TimesFM 2.5** (recommended) | 200M | ≥ 4 GB | ≥ 2 GB | ~800 MB | up to 16,384 |
| TimesFM 2.0 (archived) | 500M | ≥ 16 GB | ≥ 8 GB | ~2 GB | up to 2,048 |
| TimesFM 1.0 (archived) | 200M | ≥ 8 GB | ≥ 4 GB | ~800 MB | up to 2,048 |

> **Recommendation**: Always use TimesFM 2.5 unless you have a specific reason to use an
> older checkpoint. It is smaller, faster, and supports 8× longer context.

## 🔧 Installation

### Step 1: Verify System (always first)

```bash
python scripts/check_system.py
```

### Step 2: Install TimesFM

```bash
# Using uv (recommended by this repo)
uv pip install timesfm[torch]

# For JAX/Flax backend (faster on TPU/GPU)
uv pip install timesfm[flax]
```

### Step 3: Install PyTorch for Your Hardware

```bash
# CUDA 12.1 (NVIDIA GPU)
uv pip install torch>=2.0.0 --index-url https://download.pytorch.org/whl/cu121

# CPU only
uv pip install torch>=2.0.0 --index-url https://download.pytorch.org/whl/cpu

# Apple Silicon (MPS)
uv pip install torch>=2.0.0  # MPS support is built-in
```

### Step 4: Verify Installation

```python
import timesfm
import numpy as np
print(f"TimesFM version: {timesfm.__version__}")
print("Installation OK")
```

## 🎯 Quick Start

### Minimal Example (5 Lines)

```python
import torch, numpy as np, timesfm

torch.set_float32_matmul_precision("high")

model = timesfm.TimesFM_2p5_200M_torch.from_pretrained(
    "google/timesfm-2.5-200m-pytorch"
)
model.compile(timesfm.ForecastConfig(
    max_context=1024, max_horizon=256, normalize_inputs=True,
    use_continuous_quantile_head=True, force_flip_invariance=True,
    infer_is_positive=True, fix_quantile_crossing=True,
))

point, quantiles = model.forecast(horizon=24, inputs=[
    np.sin(np.linspace(0, 20, 200)),  # any 1-D array
])
# point.shape == (1, 24)        — median forecast
# quantiles.shape == (1, 24, 10) — 10th–90th percentile bands
```

### Forecast from CSV

```python
import pandas as pd, numpy as np

df = pd.read_csv("monthly_sales.csv", parse_dates=["date"], index_col="date")

# Convert each column to a list of arrays
inputs = [df[col].dropna().values.astype(np.float32) for col in df.columns]

point, quantiles = model.forecast(horizon=12, inputs=inputs)

# Build a results DataFrame
for i, col in enumerate(df.columns):
    last_date = df[col].dropna().index[-1]
    future_dates = pd.date_range(last_date, periods=13, freq="MS")[1:]
    forecast_df = pd.DataFrame({
        "date": future_dates,
        "forecast": point[i],
        "lower_80": quantiles[i, :, 2],  # 20th percentile
        "upper_80": quantiles[i, :, 8],  # 80th percentile
    })
    print(f"\n--- {col} ---")
    print(forecast_df.to_string(index=False))
```

### Forecast with Covariates (XReg)

TimesFM 2.5+ supports exogenous variables through `forecast_with_covariates()`. Requires `timesfm[xreg]`.

```python
# Requires: uv pip install timesfm[xreg]
point, quantiles = model.forecast_with_covariates(
    inputs=inputs,
    dynamic_numerical_covariates={"price": price_arrays},
    dynamic_categorical_covariates={"holiday": holiday_arrays},
    static_categorical_covariates={"region": region_labels},
    xreg_mode="xreg + timesfm",  # or "timesfm + xreg"
)
```

| Covariate Type | Description | Example |
| -------------- | ----------- | ------- |
| `dynamic_numerical` | Time-varying numeric | price, temperature, promotion spend |
| `dynamic_categorical` | Time-varying categorical | holiday flag, day of week |
| `static_numerical` | Per-series numeric | store size, account age |
| `static_categorical` | Per-series categorical | store type, region, product category |

**XReg Modes:**
- `"xreg + timesfm"` (default): TimesFM forecasts first, then XReg adjusts residuals
- `"timesfm + xreg"`: XReg fits first, then TimesFM forecasts residuals

> See `examples/covariates-forecasting/` for a complete example with synthetic retail data.

### Anomaly Detection (via Quantile Intervals)

TimesFM does not have built-in anomaly detection, but the **quantile forecasts naturally provide
prediction intervals** that can detect anomalies:

```python
point, q = model.forecast(horizon=H, inputs=[values])

# 90% prediction interval
lower_90 = q[0, :, 1]  # 10th percentile
upper_90 = q[0, :, 9]  # 90th percentile

# Detect anomalies: values outside the 90% CI
actual = test_values  # your holdout data
anomalies = (actual < lower_90) | (actual > upper_90)

# Severity levels
is_warning = (actual < q[0, :, 2]) | (actual > q[0, :, 8])  # outside 80% CI
is_critical = anomalies  # outside 90% CI
```

| Severity | Condition | Interpretation |
| -------- | --------- | -------------- |
| **Normal** | Inside 80% CI | Expected behavior |
| **Warning** | Outside 80% CI | Unusual but possible |
| **Critical** | Outside 90% CI | Statistically rare (< 10% probability) |

> See `examples/anomaly-detection/` for a complete example with visualization.

```python
# Requires: uv pip install timesfm[xreg]
point, quantiles = model.forecast_with_covariates(
    inputs=inputs,
    dynamic_numerical_covariates={"temperature": temp_arrays},
    dynamic_categorical_covariates={"day_of_week": dow_arrays},
    static_categorical_covariates={"region": region_labels},
    xreg_mode="xreg + timesfm",  # or "timesfm + xreg"
)
```

## Output, Configuration, Workflows, and Tuning

- [references/output_and_config.md](references/output_and_config.md): reading the point
  forecast and the 10 quantile bands, deriving prediction intervals, and every
  `ForecastConfig` field.
- [references/workflows.md](references/workflows.md): the standard forecast sequence,
  many-series forecasting from a wide CSV, and backtesting with interval coverage.
- [references/performance_tuning.md](references/performance_tuning.md): GPU and TF32
  setup, `per_core_batch_size` by available memory, and memory management.
- [references/examples_and_validation.md](references/examples_and_validation.md):
  runnable examples, the quality checklist, common mistakes, and regression checks.

## 🔗 Integration with Other Skills

### With `statsmodels`

Use `statsmodels` for classical models (ARIMA, SARIMAX) as a **comparison baseline**:

```python
# TimesFM forecast
tfm_point, tfm_q = model.forecast(horizon=H, inputs=[values])

# statsmodels ARIMA forecast
from statsmodels.tsa.arima.model import ARIMA
arima = ARIMA(values, order=(1,1,1)).fit()
arima_forecast = arima.forecast(steps=H)

# Compare
print(f"TimesFM MAE: {np.mean(np.abs(actual - tfm_point[0])):.2f}")
print(f"ARIMA MAE:   {np.mean(np.abs(actual - arima_forecast)):.2f}")
```

### With `matplotlib` / `scientific-visualization`

Plot forecasts with prediction intervals as publication-quality figures.

### With `exploratory-data-analysis`

Run EDA on the time series before forecasting to understand trends, seasonality, and stationarity.

## 📚 Available Scripts

### `scripts/check_system.py`

**Mandatory preflight checker.** Run before first model load.

```bash
python scripts/check_system.py
```

Output example:
```
=== TimesFM System Requirements Check ===

[RAM]       Total: 32.0 GB | Available: 24.3 GB  ✅ PASS
[GPU]       NVIDIA RTX 4090 | VRAM: 24.0 GB      ✅ PASS
[Disk]      Free: 142.5 GB                        ✅ PASS
[Python]    3.12.1                                 ✅ PASS
[timesfm]   Installed (2.5.0)                      ✅ PASS
[torch]     Installed (2.4.1+cu121)                ✅ PASS

VERDICT: ✅ System is ready for TimesFM 2.5 (GPU mode)
Recommended: per_core_batch_size=128
```

### `scripts/forecast_csv.py`

End-to-end CSV forecasting with automatic system check.

```bash
python scripts/forecast_csv.py input.csv \
    --horizon 24 \
    --date-col date \
    --value-cols sales,revenue \
    --output forecasts.csv
```

## 📖 Reference Documentation

Detailed guides in `references/`:

| File | Contents |
| ---- | -------- |
| `references/system_requirements.md` | Hardware tiers, GPU/CPU selection, memory estimation formulas |
| `references/api_reference.md` | Full `ForecastConfig` docs, `from_pretrained` options, output shapes |
| `references/data_preparation.md` | Input formats, NaN handling, CSV loading, covariate setup |

## Common Pitfalls

1. **Not running system check** → model load crashes on low-RAM machines. Always run `check_system.py` first.
2. **Forgetting `model.compile()`** → `RuntimeError: Model is not compiled`. Must call `compile()` before `forecast()`.
3. **Not setting `normalize_inputs=True`** → unstable forecasts for series with large values.
4. **Using v1/v2 on machines with < 32 GB RAM** → use TimesFM 2.5 (200M params) instead.
5. **Not setting `fix_quantile_crossing=True`** → quantiles may not be monotonic (q10 > q50).
6. **Huge `per_core_batch_size` on small GPU** → CUDA OOM. Start small, increase.
7. **Passing 2-D arrays** → TimesFM expects a **list of 1-D arrays**, not a 2-D matrix.
8. **Forgetting `torch.set_float32_matmul_precision("high")`** → slower inference on Ampere+ GPUs.
9. **Not handling NaN in output** → edge cases with very short series. Always check `np.isnan(point).any()`.
10. **Using `infer_is_positive=True` for series that can be negative** → clamps forecasts at zero. Set False for temperature, returns, etc.

## Model Versions

```mermaid
timeline
    accTitle: TimesFM Version History
    accDescr: Timeline of TimesFM model releases showing parameter counts and key improvements.

    section 2024
        TimesFM 1.0 : 200M params, 2K context, JAX only
        TimesFM 2.0 : 500M params, 2K context, PyTorch + JAX
    section 2025
        TimesFM 2.5 : 200M params, 16K context, quantile head, no frequency indicator
```

| Version | Params | Context | Quantile Head | Frequency Flag | Status |
| ------- | ------ | ------- | ------------- | -------------- | ------ |
| **2.5** | 200M | 16,384 | ✅ Continuous (30M) | ❌ Removed | **Latest** |
| 2.0 | 500M | 2,048 | ✅ Fixed buckets | ✅ Required | Archived |
| 1.0 | 200M | 2,048 | ✅ Fixed buckets | ✅ Required | Archived |

**Hugging Face checkpoints:**

- `google/timesfm-2.5-200m-pytorch` (recommended)
- `google/timesfm-2.5-200m-flax`
- `google/timesfm-2.0-500m-pytorch` (archived)
- `google/timesfm-1.0-200m-pytorch` (archived)

## Resources

- **Paper**: [A Decoder-Only Foundation Model for Time-Series Forecasting](https://arxiv.org/abs/2310.10688) (ICML 2024)
- **Repository**: https://github.com/google-research/timesfm
- **Hugging Face**: https://huggingface.co/collections/google/timesfm-release-66e4be5fdb56e960c1e482a6
- **Google Blog**: https://research.google/blog/a-decoder-only-foundation-model-for-time-series-forecasting/
- **BigQuery Integration**: https://cloud.google.com/bigquery/docs/timesfm-model

## Dónde encaja

- Categoría: [Datos y analítica](https://skillsagentes.com/categorias/datos-analitica.md) — Consulta, limpia y visualiza datos sin salir del agente.
- Creador: [K-Dense-AI](https://skillsagentes.com/creators/k-dense-ai.md) — 163 skills en el directorio
- [Todas las skills](https://skillsagentes.com/skills.md)
- [Ranking de instalaciones](https://skillsagentes.com/ranking.md)

## Otras skills del mismo repositorio

- [Citation Management](https://skillsagentes.com/skills/k-dense-ai/scientific-agent-skills/citation-management.md): Gestión integral de citas académicas: busca en OpenAlex, PubMed y Google Scholar, extrae metadatos precisos, valida citas y genera entradas BibTeX correctamente formateadas.
- [Scientific Slides](https://skillsagentes.com/skills/k-dense-ai/scientific-agent-skills/scientific-slides.md): Crea decks de diapositivas y presentaciones para charlas de investigación: PowerPoint, presentaciones de conferencia, seminarios, defensas de tesis. Da estructura, plantillas, guía de tiempos y validación visual.
- [Literature Review](https://skillsagentes.com/skills/k-dense-ai/scientific-agent-skills/literature-review.md): Realiza revisiones bibliográficas sistemáticas y completas usando varias bases académicas (PubMed, arXiv, bioRxiv, Semantic Scholar). Genera markdown y PDF con citas verificadas en varios estilos (APA, Nature, Vancouver).
- [Infographics](https://skillsagentes.com/skills/k-dense-ai/scientific-agent-skills/infographics.md): Crea infografías profesionales con Nano Banana Pro AI y refinamiento iterativo inteligente. Usa Gemini 3.6 Flash para revisar la calidad e integra investigación con Perplexity Sonar. Soporta 10 tipos, 8 estilos y paletas para daltonismo.
- [Latex Posters](https://skillsagentes.com/skills/k-dense-ai/scientific-agent-skills/latex-posters.md): Crea pósteres de investigación profesionales en LaTeX con beamerposter, tikzposter o baposter, para conferencias y comunicación científica: layout, colores, columnas múltiples e integración de figuras.

---

Skills Agentes · [Índice de páginas en markdown](https://skillsagentes.com/sitemap.md) · [Inicio](https://skillsagentes.com/index.md)