theforecastingcompany/GiftEvalPlots

---
dataset_info:
- config_name: default
features:
- name: dataset_name
dtype: string
- name: item_id
dtype: string
- name: start
dtype: timestamp[s]
- name: freq
dtype: string
- name: target
list:
list: float32
- name: past_feat_dynamic_real
list:
list: float32
splits:
- name: valid
num_examples: 26
configs:
- config_name: default
default: true
data_files:
- split: valid
path: data/valid-*
license: cc-by-4.0
task_categories:
- time-series-forecasting
tags:
- time-series
- forecasting
- gift-eval
- evaluation
---

# GiftEvalPlots

A curated subset of [theforecastingcompany/GiftEval](https://huggingface.co/datasets/theforecastingcompany/GiftEval) containing 26 representative time series selected for forecast visualization and model comparison.

## Usage

```python
from datasets import load_dataset

ds = load_dataset("theforecastingcompany/GiftEvalPlots", split="valid")
```

## Contents

This dataset contains 26 time series from 11 GiftEval subdatasets, spanning multiple frequencies and domains:

| Source Dataset | Frequency | item_ids | Count |
|----------------|-----------|----------|-------|
| `bizitobs_application` | 10S | Robot-Shop | 1 |
| `bizitobs_l2c` | H | L2C | 1 |
| `bizitobs_service` | 10S | discount, nginx-web, payment | 3 |
| `car_parts_with_missing` | M | T2618, T2622, T2670, T2673 | 4 |
| `covid_deaths` | D | T1, T150 | 2 |
| `electricity` | W | MT_026, MT_228 | 2 |
| `hierarchical_sales` | D | QTY_B1_12, QTY_B1_37, QTY_B4_5 | 3 |
| `m4_quarterly` | Q-DEC | 331, 15689, 23660 | 3 |
| `restaurant` | D | air_258dc112912fc458, air_32460819c7600037, air_8c119d1f16049f20, air_93b9bb641f8fc982, air_ba937bf13d40fb24 | 5 |
| `us_births` | M | T1 | 1 |
| `us_births` | W | T1 | 1 |

## Schema

| Column | Type | Description |
|--------|------|-------------|
| `dataset_name` | string | Source GiftEval subdataset name |
| `item_id` | string | Unique time series identifier within the subdataset |
| `start` | timestamp[s] | Start timestamp of the time series |
| `freq` | string | Pandas-compatible frequency string |
| `target` | list[list[float32]] | Target values (list of variates, each a list of floats) |
| `past_feat_dynamic_real` | list[list[float32]] | Past dynamic real-valued features |

## Source

Extracted from [theforecastingcompany/GiftEval](https://huggingface.co/datasets/theforecastingcompany/GiftEval), which is a Parquet conversion of [Salesforce/GiftEval](https://huggingface.co/datasets/Salesforce/GiftEval).