rogerdehe commited on
Commit
adea16c
·
verified ·
1 Parent(s): 8eb7b3d

add dataset card

Browse files
Files changed (1) hide show
  1. README.md +82 -0
README.md ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: mit
3
+ task_categories:
4
+ - time-series-forecasting
5
+ language:
6
+ - en
7
+ tags:
8
+ - crypto
9
+ - cryptocurrency
10
+ - finance
11
+ - ohlcv
12
+ - klines
13
+ - gateio
14
+ - trading
15
+ pretty_name: Gate.io USDT OHLCV Klines
16
+ size_categories:
17
+ - 1B<n<10B
18
+ ---
19
+
20
+ # Gate.io USDT OHLCV Klines
21
+
22
+ Historical OHLCV candlestick (kline) data for **Gate.io** USDT-quoted markets,
23
+ stored as zstd-compressed Parquet.
24
+
25
+ ## Coverage
26
+
27
+ | | |
28
+ |---|---|
29
+ | **Exchange** | Gate.io |
30
+ | **Markets** | `spot` (primary) · `futures` (USDT perpetuals — limited history, see Notes) |
31
+ | **Timeframes** | 5m · 15m · 1h · 4h · 1d (all pairs) — plus 1m for a ~1.3k-pair subset |
32
+ | **Pairs** | ~3,300 spot USDT pairs |
33
+ | **History** | spot back to **2013** for the oldest listings |
34
+ | **As of** | static snapshot (~2025-05); periodic refresh planned |
35
+ | **Source** | seeded from the DigiTuccar `HistoricalDataForTradeBacktest` snapshot |
36
+
37
+ ## File layout
38
+
39
+ ```
40
+ {market}/{timeframe}/{PAIR}.parquet
41
+ ```
42
+
43
+ Examples: `spot/1d/BTC_USDT.parquet`, `spot/5m/ETH_USDT.parquet`
44
+
45
+ Each Parquet holds **one pair × one timeframe**. Columns:
46
+
47
+ | column | type | description |
48
+ |---|---|---|
49
+ | `date` | datetime (UTC) | candle open time |
50
+ | `open` | float | open price |
51
+ | `high` | float | high price |
52
+ | `low` | float | low price |
53
+ | `close` | float | close price |
54
+ | `volume` | float | base-asset volume |
55
+
56
+ ## Usage
57
+
58
+ Single file:
59
+ ```python
60
+ from huggingface_hub import hf_hub_download
61
+ import pandas as pd
62
+
63
+ path = hf_hub_download("rogerdehe/klines-gateio",
64
+ "spot/1d/BTC_USDT.parquet", repo_type="dataset")
65
+ df = pd.read_parquet(path)
66
+ ```
67
+
68
+ A subset (or all) of the dataset:
69
+ ```python
70
+ from huggingface_hub import snapshot_download
71
+ snapshot_download("rogerdehe/klines-gateio", repo_type="dataset",
72
+ allow_patterns=["spot/1d/*"]) # e.g. just daily spot
73
+ ```
74
+
75
+ ## Notes
76
+
77
+ - **Futures history is limited**: Gate.io's API only served ~180 days back at
78
+ snapshot time, so `futures/` is thin relative to `spot/`. Use spot for deep history.
79
+ - `1m` is available only for a subset of pairs (~1.3k); other timeframes cover all pairs.
80
+ - The HF **dataset viewer is disabled** (custom per-pair layout); load files directly as shown above.
81
+ - Timestamps are UTC. Data is provided **as-is** for research purposes, with no
82
+ warranty of accuracy or completeness.