improvements

docs/source/index.md

@@ -1,7 +1,7 @@
 MolFlux
 =======
 
-**MolFlux** is a foundational package for molecular property prediction.
+**MolFlux** is a foundational package for molecular predictive modelling.
 
 ```{note}
    These docs are under active dev

docs/source/pages/datasets/intro.md

@@ -10,4 +10,6 @@ The ``datasets`` submodule is meant to address all of these issues. It is a libr
 multiple sources. Whether you are looking for public datasets (such as PDBBind or QM9) or just easy access to saved data,
 ``datasets`` provides a standard and modular interface for accessing and manipulating these datasets!
 
-It is built on top of [huggingface ``datasets``](https://huggingface.co/docs/datasets/index).
+It is built on top of [huggingface ``datasets``](https://huggingface.co/docs/datasets/index). The huggingface ``datasets``
+package is versatile, fast, and efficient. It can handle many data types and its built-in functionality allows for scalable
+and fast data manipulation and handling.

docs/source/pages/modelzoo/basic_usage.md

@@ -43,8 +43,8 @@ This returns our catalogue of available model architectures (organised by the de
 ```{seealso}
 [How to add your own model](how_to_add_models.md) if you would like to add your own model to the catalogue
 ```
- For instance, `molflux.modelzoo.list_models()` returns as one item in the dictionary: 
-`'xgboost': ['xg_boost_classifier', 'xg_boost_regressor']`. In order to be able to use the two models `xg_boost_classifier` 
+ For instance, `molflux.modelzoo.list_models()` returns as one item in the dictionary:
+`'xgboost': ['xg_boost_classifier', 'xg_boost_regressor']`. In order to be able to use the two models `xg_boost_classifier`
 and `xg_boost_regressor`, you would do: ``pip install molflux[xgboost]``.
 
 ## Loading a model architecture
@@ -287,8 +287,12 @@ predictions = model.predict(test_data)
 print(predictions)
 ```
 
-This returns a dictionary of your model's predictions!
+This returns a dictionary of your model's predictions! Models can also support different inference methods. For example,
+some classification models support the ``predict_proba`` method which returns the probabilities of the classes
 
+```python
+probabilities = model.predict_proba(test_data)
+```
 
 ## Saving/Loading a model
 
@@ -307,7 +311,9 @@ save_to_store("path_to_my_model/", model)
 The ``save_to_store`` function takes the path and the model to save. It can save to local disk or to an s3 location.
 
 ```{note}
-Recommend using prod saving ..., link to core
+For models intended for production level usage, we recommend that they are saved as described in the [productionising](../production/models.md)
+section. Along with the model, this also save the featurisation metadata and a snapshot of the environment the model was
+built in.
 ```
 
 ### Loading

docs/source/pages/modelzoo/uncertainty.md

@@ -42,7 +42,7 @@ models implement additional functionalities:
     while their spread indicates how uncertain the model is about this input.
   4) `calibrate_uncertainty(data, **kwargs)` - calibrates the uncertainty of this model to an external/validation dataset
 
-You can check whether a model implements any of these methods by using the appropriate `supports_x()` utility function:
+You can check whether a model implements any of these methods by using the appropriate `supports_*` utility function:
 
 ```{code-cell} ipython3
 from molflux.modelzoo import load_model, supports_prediction_interval
@@ -61,8 +61,8 @@ Similarly, `supports_std`, `supports_sampling`, and `supports_uncertainty_calibr
 
 ### Quick example - CatBoost Models
 
-The typical example for a model with implemented uncertainty methods is a CatBoost Model.
-Gaussian processes are distributions over functions such that predictions can be characterized by a mean and covariance.
+A typical example for a model with implemented uncertainty methods is the CatBoost Model. This model architecture can
+return both a mean and standard deviation prediction.
 
 In the example below, we will train and predict using a CatBoost, and then use some of the
 functions defined above to get a measure of the model uncertainty.
@@ -113,7 +113,7 @@ fitted `k` times.
 2. In two steps - first, training an underlying model on training data, then
 calibrating the uncertainty of it on a validation dataset
 
-Both of these are possible with our Mapie implementation.
+Both of these are possible with our [Mapie](https://github.com/scikit-learn-contrib/MAPIE) implementation.
 
 ```{note}
 This functionality is still work in progress.

docs/source/pages/splits/gallery.md

@@ -32,7 +32,7 @@ The original `sklearn` [notebook](https://scikit-learn.org/stable/auto_examples/
 
 ```{code-cell} ipython3
 ---
-tags: [remove-input]
+tags: [hide-input]
 ---
 import numpy as np
 import matplotlib.pyplot as plt
@@ -59,7 +59,7 @@ To begin, we'll visualize our data:
 
 ```{code-cell} ipython3
 ---
-tags: [remove-input]
+tags: [hide-input]
 ---
 # Generate the class/group data
 n_points = 100
@@ -111,7 +111,7 @@ the validation set (in grey), and the test set (in red).
 
 ```{code-cell} ipython3
 ---
-tags: [remove-input]
+tags: [hide-input]
 ---
 def plot_cv_indices(cv, X, y, group, ax, n_splits, lw=10):
     """Create a sample plot for indices of a cross-validation object."""
@@ -170,7 +170,7 @@ Let's see how it looks for the `k_fold` cross-validation object:
 
 ```{code-cell} ipython3
 ---
-tags: [remove-input]
+tags: [hide-input]
 ---
 fig, ax = plt.subplots(figsize=figsize)
 strategy = load_splitting_strategy("k_fold")
@@ -186,7 +186,7 @@ consideration. We can change this by using either:
 
 ```{code-cell} ipython3
 ---
-tags: [remove-input]
+tags: [hide-input]
 ---
 strategies = ["stratified_k_fold", "group_k_fold"]
 
@@ -207,7 +207,7 @@ Note how some use the group/class information while others do not:
 
 ```{code-cell} ipython3
 ---
-tags: [remove-input]
+tags: [hide-input]
 ---
 strategies = ["group_k_fold", "group_shuffle_split", "k_fold", "linear_split", "shuffle_split", "stratified_k_fold", "stratified_shuffle_split", "time_series_split"]
 

docs/source/pages/tutorials/esol_training.md

@@ -31,6 +31,7 @@ print(dataset)
 print(dataset[0])
 ```
 
+The loaded dataset is an instance of a HuggingFace ``Dataset`` (for more info, checkout the [docs](https://huggingface.co/docs/datasets/index)).
 You can see that there are two columns: ``smiles`` and ``log_solubility``.
 
 
@@ -127,6 +128,7 @@ plt.scatter(
     split_featurised_dataset["test"]["log_solubility"],
     preds["random_forest_regressor::log_solubility"],
 )
+plt.plot([-10, 0], [-10, 0], c='r')
 plt.xlabel("True values")
 plt.ylabel("Predicted values")
 plt.show()
@@ -234,6 +236,7 @@ plt.scatter(
     split_featurised_dataset["test"]["log_solubility"],
     preds["random_forest_regressor::log_solubility"],
 )
+plt.plot([-10, 0], [-10, 0], c='r')
 plt.xlabel("True values")
 plt.ylabel("Predicted values")
 plt.show()

src/molflux/datasets/builders/ani2x/ani2x.py

@@ -7,7 +7,7 @@
 import datasets
 from molflux.datasets.typing import ExamplesGenerator
 
-from .ani2x_configs import FEATURES, LEVEL_OF_THEORY, URL_DICT
+from .ani2x_configs import FEATURES, LEVELS_OF_THEORY, URL_DICT
 
 _BASE_URL = "https://zenodo.org/records/10108942"
 
@@ -26,7 +26,7 @@
 @dataclass
 class ANI2XConfig(datasets.BuilderConfig):
     backend: Literal["openeye", "rdkit"] = "rdkit"
-    level_of_theory: LEVEL_OF_THEORY = "wB97X/631Gd"
+    level_of_theory: LEVELS_OF_THEORY = "wB97X/631Gd"
 
 
 class ANI2X(datasets.GeneratorBasedBuilder):

src/molflux/datasets/builders/ani2x/ani2x_configs.py

@@ -25,7 +25,7 @@
     },
 }
 
-LEVEL_OF_THEORY = Literal[
+LEVELS_OF_THEORY = Literal[
     "wB97X/631Gd",
     "wB97X/def2TZVPP",
     "wB97MD3BJ/def2TZVPP",

tests/datasets/plugins/openeye/ani2x/test_ani2x.py

@@ -6,6 +6,7 @@
 import datasets
 from molflux.datasets import load_dataset
 from molflux.datasets.builders.ani2x.ani2x import ANI2X
+from molflux.datasets.builders.ani2x.ani2x_configs import FEATURES
 from molflux.datasets.catalogue import list_datasets
 
 dataset_name = "ani2x"
@@ -87,6 +88,7 @@ def test_dataset_has_correct_num_rows(level_of_theory):
 
     dataset = load_dataset(dataset_name, backend_name, level_of_theory=level_of_theory)
 
+    assert set(dataset.column_names) == set(FEATURES[level_of_theory].keys())
     assert len(dataset) == 10
 
 
@@ -95,8 +97,8 @@ def test_dataset_has_correct_num_rows(level_of_theory):
     levels_of_theory,
 )
 @pytest.mark.usefixtures("_fixture_mocked_dataset_asset")
-def test_dataset_is_readable_with_rdkit(level_of_theory):
-    """That rdkit can read the mol bytes"""
+def test_dataset_is_readable_with_openeye(level_of_theory):
+    """That openeye can read the mol bytes"""
 
     dataset = load_dataset(dataset_name, backend_name, level_of_theory=level_of_theory)
 

tests/datasets/plugins/rdkit/ani2x/test_ani2x.py

@@ -6,6 +6,7 @@
 import datasets
 from molflux.datasets import load_dataset
 from molflux.datasets.builders.ani2x.ani2x import ANI2X
+from molflux.datasets.builders.ani2x.ani2x_configs import FEATURES
 from molflux.datasets.catalogue import list_datasets
 
 dataset_name = "ani2x"
@@ -86,6 +87,7 @@ def test_dataset_has_correct_num_rows(level_of_theory):
     """That the built dataset has correct num rows."""
 
     dataset = load_dataset(dataset_name, backend_name, level_of_theory=level_of_theory)
+    assert set(dataset.column_names) == set(FEATURES[level_of_theory].keys())
 
     assert len(dataset) == 10