Datasets:
LeoTungAnh
commited on
Commit
•
d834455
1
Parent(s):
9ede10d
Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: openrail
|
3 |
+
task_categories:
|
4 |
+
- time-series-forecasting
|
5 |
+
size_categories:
|
6 |
+
- 1K<n<10K
|
7 |
+
---
|
8 |
+
**Download the Dataset**:
|
9 |
+
```python
|
10 |
+
from datasets import load_dataset
|
11 |
+
|
12 |
+
dataset = load_dataset("LeoTungAnh/electricity_hourly")
|
13 |
+
```
|
14 |
+
|
15 |
+
**Dataset Card for Electricity Consumption**
|
16 |
+
|
17 |
+
This dataset encompasses hourly electricity consumption in kilowatts (kW) across a span of three years (2012-2014), involving 370 individual clients in Portugal.
|
18 |
+
|
19 |
+
**Preprocessing information**:
|
20 |
+
- Grouped by hour (frequency: "1H").
|
21 |
+
- Applied Standardization as preprocessing technique ("Std").
|
22 |
+
|
23 |
+
**Dataset information**:
|
24 |
+
- Number of time series: 370
|
25 |
+
- Number of training samples: 26208
|
26 |
+
- Number of validation samples: 26256 (number_of_training_samples + 48)
|
27 |
+
- Number of testing samples: 26304 (number_of_validation_samples + 48)
|
28 |
+
|
29 |
+
**Dataset format**:
|
30 |
+
```python
|
31 |
+
Dataset({
|
32 |
+
|
33 |
+
features: ['start', 'target', 'feat_static_cat', 'feat_dynamic_real', 'item_id'],
|
34 |
+
|
35 |
+
num_rows: 370
|
36 |
+
|
37 |
+
})
|
38 |
+
```
|
39 |
+
**Data format for a sample**:
|
40 |
+
|
41 |
+
- 'start': datetime.datetime
|
42 |
+
|
43 |
+
- 'target': list of a time series data
|
44 |
+
|
45 |
+
- 'feat_static_cat': time series index
|
46 |
+
|
47 |
+
- 'feat_dynamic_real': None
|
48 |
+
|
49 |
+
- 'item_id': name of time series
|
50 |
+
|
51 |
+
|
52 |
+
**Data example**:
|
53 |
+
```python
|
54 |
+
{'start': datetime.datetime(2012, 1, 1, 1, 0),
|
55 |
+
|
56 |
+
'target': [-0.19363673541224083, -0.08851588245610625, -0.19363673541224083, ... -0.5615597207587115,...],
|
57 |
+
|
58 |
+
'feat_static_cat': [0],
|
59 |
+
|
60 |
+
'feat_dynamic_real': None,
|
61 |
+
|
62 |
+
'item_id': 'MT_001'
|
63 |
+
}
|
64 |
+
```
|
65 |
+
|
66 |
+
**Usage**:
|
67 |
+
- The dataset can be used by available Transformer, Autoformer, Informer of Huggingface.
|
68 |
+
- Other algorithms can extract data directly by making use of 'target' feature.
|