Wrap lightning's train as an ST method (#40)

* add train method to starling class, injecting defaults

* return AnnData object from result method

* remove redundant type descriptions in docstrings

* update second tutorial link in README

* remove notebook output

* only check for size column if model_cell_size is True

Author: cklamann

Diff for: Dockerfile

@@ -35,6 +35,6 @@ RUN python3 -m venv $VIRTUAL_ENV && \
 
 USER $USERNAME
 
-COPY . .
+COPY --chown=${USER_UID}:${USER_GID} . .
 
 RUN poetry install --with docs,dev

Diff for: README.md

@@ -11,7 +11,7 @@ It outputs:
 
 A **preprint** describing the method and introducing a novel benchmarking workflow is available: [Lee et al. (2024) _Segmentation error aware clustering for highly multiplexed imaging_](https://www.biorxiv.org/content/10.1101/2024.02.29.582827v1)
 
-A **tutorial** outlining basic usage is available [here](https://github.com/camlab-bioml/starling/blob/main/docs/source/tutorial/getting-started.ipynb).
+A **tutorial** outlining basic usage is available [here][tutorial].
 
 ![Model](https://github.com/camlab-bioml/starling/raw/main/starling-schematic600x.png)
 

Diff for: docs/source/conf.py

@@ -21,6 +21,8 @@
 
 nb_execution_timeout = -1
 
+autodoc_typehints = "description"
+
 html_theme = "sphinx_rtd_theme"
 html_static_path = ["_static"]
 html_css_files = [