# Market Lab v0.1.6 (/changelog/v0.1.6)



> v0.1.6 adds a JupyterLab research mode to Market Lab.

Run one command inside an existing Python project and Market Lab launches JupyterLab with historical data and Rust-backed studies available through a preloaded `mlab` object.

No Market Lab Python package is required. The notebook continues to use the project’s own Python environment, dependencies, and research workflow.

## Launch from your Python project [#start-from-the-cli]

From the project directory, run:

```bash
mlab notebook
```

Market Lab opens JupyterLab in the current folder.

You can select an interpreter with `--python`. Otherwise, Market Lab can use the project’s local `.venv` or the Python available on `PATH`.

The selected environment retains its own NumPy, pandas, Matplotlib, and other research packages. Market Lab does not replace the project’s dependency setup with a separate notebook environment.

If JupyterLab or `ipykernel` is unavailable, the CLI prints the appropriate installation command for the project’s package manager.

## Market Lab inside the kernel [#research-cell-by-cell]

Every notebook launched through `mlab notebook` receives a ready-to-use `mlab` object.

Create a historical range and request data directly from a cell:

```python
history = mlab.history(
    start="2026-08-01",
    end="2026-08-20",
)

candles = history.source(
    "btc@candles@hyperliquidf:timeframe=3600"
)
```

Historical sources are loaded lazily.

The first call fetches and normalizes the requested range. Repeated calls for the same source use the notebook’s in-memory cache instead of fetching the data again.

Returned records are ordinary Python dictionaries, so they can be passed directly into pandas, NumPy, or the rest of the project’s existing research stack.

## Rust-backed studies [#native-studies]

Notebook cells can call Market Lab’s native Rust study engine directly:

```python
sma = mlab.study.sma(
    candles,
    {
        "field": "c",
        "window": 20,
    },
)
```

Checkout [Built in functions](/scripting-v2/built-ins) to see supported studies

## Designed for research, not simulation [#research-only-boundary]

Notebook is meant for research purpose only

It does not expose: `ctx`, Strategy hooks, The Market Lab simulator, Positions or Live execution

Use Scripting V2 with plain python files:

```bash
mlab script backtest
```

when a strategy must process data sequentially through the Market Lab simulator.

Use:

```bash
mlab script run
```

when the same strategy is ready for live daemon deployment.

See [Jupyter Notebook](/scripting-v2/notebook) for environment setup, the cell-by-cell workflow, source caching, and the local bridge design.