Update README.md

README.md

@@ -28,66 +28,50 @@ configs:
     path: data/ohlcv_*.parquet
 ---
 
+# 📈 OHLCV-1m: US Stock Market Minute-Level Candlestick Data (1992–2025)
 
-# OHLCV-1m Dataset
+This dataset provides minute-level OHLCV (Open, High, Low, Close, Volume) candlestick data for thousands of U.S. stocks across multiple decades (1992 to 2025). The data was originally sourced from [Finnhub.io](https://finnhub.io), a real-time market data provider.
 
-This dataset contains 1-minute OHLCV data by month, from 1992 onward.
+It has been aggregated and reformatted from monthly `.tar` archives into clean and unified Parquet files — one per month — and uploaded to the Hugging Face Hub for easy access.
 
-Each monthly file is stored as a Parquet file in the `data/` folder.
+## 🧾 Dataset Structure
 
-## How to Load a File
+Each row in the dataset represents **one minute** of trading for a given stock ticker, and includes the following columns:
+
+| Column     | Type                         | Description                         |
+|------------|------------------------------|-------------------------------------|
+| `timestamp`| `datetime64[ns, UTC]`        | Start time of the minute            |
+| `open`     | `float64`                    | Opening price                       |
+| `high`     | `float64`                    | Highest price within the minute     |
+| `low`      | `float64`                    | Lowest price within the minute      |
+| `close`    | `float64`                    | Closing price                       |
+| `volume`   | `float64`                    | Volume traded within the minute     |
+| `ticker`   | `string`                     | Stock ticker symbol                 |
+
+The data is split by month into files like:
+
+data/ohlcv_1992-01.parquet
+data/ohlcv_1992-02.parquet
+...
+data/ohlcv_2025-05.parquet
+
+
+## 📚 Usage
 
 ```python
-import pandas as pd
-df = pd.read_parquet("https://huggingface.co/datasets/mito0o852/OHLCV-1m/resolve/main/data/ohlcv_1992-01.parquet")
+from datasets import load_dataset
 
+# Load the dataset (will stream across all months)
+ds = load_dataset("mito0o852/OHLCV-1m", split="train")
 
+# View one row
+print(ds[0])
 
-This doesn't fix the viewer, but it makes it usable for others.
 
----
 
-### ✅ Solution 3: Use a `dataset loading script`
+# To convert it into a pandas DataFrame:
 
-To **enable the Hugging Face Dataset Viewer**, you need to add a script in the repo like:
+import pandas as pd
 
-```python
-# ohclv_loader.py
-import datasets
-
-class OHLCVDataset(datasets.GeneratorBasedBuilder):
-    def _info(self):
-        return datasets.DatasetInfo(
-            features=datasets.Features({
-                'timestamp': datasets.Value('timestamp[s]'),
-                'open': datasets.Value('float32'),
-                'high': datasets.Value('float32'),
-                'low': datasets.Value('float32'),
-                'close': datasets.Value('float32'),
-                'volume': datasets.Value('float32'),
-                'ticker': datasets.Value('string')
-            })
-        )
-
-    def _split_generators(self, dl_manager):
-        parquet_files = [f"data/ohlcv_{year}-{str(month).zfill(2)}.parquet"
-                         for year in range(1992, 2025+1)
-                         for month in range(1, 13)]
-        urls = [f"https://huggingface.co/datasets/mito0o852/OHLCV-1m/resolve/main/{f}" for f in parquet_files]
-        return [datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"files": dl_manager.download(urls)})]
-
-    def _generate_examples(self, files):
-        key = 0
-        for file in files:
-            df = pd.read_parquet(file)
-            for row in df.itertuples(index=False):
-                yield key, {
-                    'timestamp': row.timestamp,
-                    'open': row.open,
-                    'high': row.high,
-                    'low': row.low,
-                    'close': row.close,
-                    'volume': row.volume,
-                    'ticker': row.ticker,
-                }
-                key += 1
+df = ds.to_pandas()
+print(df.head())