fix logo in documentation (#111)

Fixes #106

docs/source/conf.py

@@ -24,10 +24,10 @@
 ]
 
 html_theme = 'pyg_sphinx_theme'
-html_logo = ('https://github.com/pyg-team/pytorch-frame/'
-             'blob/master/docs/source/_figures/pytorch_frame_logo.JPG')
+html_logo = ('https://raw.githubusercontent.com/pyg-team/pyg_sphinx_theme/'
+             'master/pyg_sphinx_theme/static/img/pytorch_frame_logo.png')
 html_favicon = ('https://raw.githubusercontent.com/pyg-team/pyg_sphinx_theme/'
-                'master/pyg_sphinx_theme/static/img/favicon.png')
+                'master/pyg_sphinx_theme/static/img/pytorch_frame_favicon.png')
 html_static_path = ['_static']
 templates_path = ['_templates']
 

docs/source/get_started/introduction.rst

@@ -1,27 +1,27 @@
 Introduction by Example
 =======================
 
-:pyg:`PyTorch Frame` is a tabular deep learning extension library for :pytorch:`null` `PyTorch <https://pytorch.org>`_.
+:pyf:`PyTorch Frame` is a tabular deep learning extension library for :pytorch:`null` `PyTorch <https://pytorch.org>`_.
 Modern data is stored in a table format with heterogeneous columns with different semantic types, e.g., numerical (e.g., age, price), categorical (e.g., gender, product type), time, texts (e.g., descriptions), images (e.g., pictures) etc.
 The goal of Pytorch Frame is to build a deep learning framework to perform effective machine learning on such a complex data.
 
 Many recent tabular models follow the modular design of :obj:`FeatureEncoder`, :obj:`TableConv`, and :obj:`Decoder`.
-:pyg:`PyTorch Frame` is designed to facilitate the creation, implementation and evaluation of deep learning models for tabular data under such modular architecture.
+:pyf:`PyTorch Frame` is designed to facilitate the creation, implementation and evaluation of deep learning models for tabular data under such modular architecture.
 Please refer to the :doc:`/get_started/modular_design` page for more information.
 
-In this doc, we shortly introduce the fundamental concepts of :pyg:`PyTorch Frame` through self-contained examples.
+In this doc, we shortly introduce the fundamental concepts of :pyf:`PyTorch Frame` through self-contained examples.
 
-At its core, :pyg:`PyTorch Frame` provides the following main features:
+At its core, :pyf:`PyTorch Frame` provides the following main features:
 
 .. contents::
     :local:
 
 Common Benchmark Datasets
 -------------------------
-:pyg:`PyTorch Frame` contains a large number of common benchmark datasets, *e.g.*, datasets from `https://github.com/yandex-research/tabular-dl-revisiting-models <https://github.com/yandex-research/tabular-dl-revisiting-models>`_
+:pyf:`PyTorch Frame` contains a large number of common benchmark datasets, *e.g.*, datasets from `https://github.com/yandex-research/tabular-dl-revisiting-models <https://github.com/yandex-research/tabular-dl-revisiting-models>`_
 , datasets from `tabular benchmark <https://huggingface.co/datasets/inria-soda/tabular-benchmark>`_ .
 
-Initializing datasets is straightforward in :pyg:`PyTorch Frame`.
+Initializing datasets is straightforward in :pyf:`PyTorch Frame`.
 An initialization of a dataset will automatically download its raw files and process the columns, *e.g*., to load the `Adult Census Income` dataset, type:
 
 .. code-block:: python
@@ -53,13 +53,13 @@ An initialization of a dataset will automatically download its raw files and pro
 Data Handling of Tables
 -----------------------
 A table contains different columns with different data types. Each data type is described by a semantic type which we refer to as :class:`~torch_frame.stype`.
-Currently :pyg:`PyTorch Frame` supports the following :class:`~torch_frame.stype`'s:
+Currently :pyf:`PyTorch Frame` supports the following :class:`~torch_frame.stype`'s:
 
 - :class:`~torch_frame.stype.categorical` denotes categorical columns.
 - :obj:`~torch_frame.stype.numerical` denotes numerical columns.
 - :obj:`~torch_frame.stype.text_embedded` denotes text columns that are pre-embedded via some sentence encoder.
 
-A table in :pyg:`PyTorch Frame` is described by an instance of :class:`~torch_frame.TensorFrame`, which holds the following attributes by default:
+A table in :pyf:`PyTorch Frame` is described by an instance of :class:`~torch_frame.TensorFrame`, which holds the following attributes by default:
 
 - :obj:`col_names_dict`: A dictionary holding the column names for each :class:`~torch_frame.stype`.
 - :obj:`feat_dict`: A dictionary holding the :obj:`Tensor` of different :class:`~torch_frame.stype`'s.
@@ -160,7 +160,7 @@ For numerical features,
 
 Mini-batches
 ------------
-Neural networks are usually trained in a mini-batch fashion. :pyg:`PyTorch Frame` contains its own :class:`torch_frame.data.DataLoader`, which can load :class:`torch_frame.data.Dataset` or :class:`~torch_frame.TensorFrame` in mini batches.
+Neural networks are usually trained in a mini-batch fashion. :pyf:`PyTorch Frame` contains its own :class:`torch_frame.data.DataLoader`, which can load :class:`torch_frame.data.Dataset` or :class:`~torch_frame.TensorFrame` in mini batches.
 
 .. code-block:: python
 
@@ -183,7 +183,7 @@ Neural networks are usually trained in a mini-batch fashion. :pyg:`PyTorch Frame
 Learning Methods on Tabular Data
 --------------------------------
 
-After learning about data handling, datasets and loader in :pyg:`PyTorch Frame`, it’s time to implement our first model!
+After learning about data handling, datasets and loader in :pyf:`PyTorch Frame`, it’s time to implement our first model!
 
 .. code-block:: python
 

docs/source/index.rst

@@ -2,7 +2,7 @@
 
 PyTorch Frame Documentation
 ===========================
-:pyg:`null` **PyTorch Frame** is a library built upon :pytorch:`null` `PyTorch <https://pytorch.org>`_ to easily write and train tabular deep learning models.
+:pyf:`null` **PyTorch Frame** is a library built upon :pytorch:`null` `PyTorch <https://pytorch.org>`_ to easily write and train tabular deep learning models.
 
 .. slack_button::
 

docs/source/modules/transforms.rst

@@ -9,7 +9,7 @@ torch_frame.transforms
 Transforms
 --------
 
-:pyg:`PyTorch Frame` allows for data transformation across different :obj:`stype`'s or within the same :obj:`stype`. Transforms takes in both :obj:`TensorFrame` and column stats.
+:pyf:`PyTorch Frame` allows for data transformation across different :obj:`stype`'s or within the same :obj:`stype`. Transforms takes in both :obj:`TensorFrame` and column stats.
 
 Let's look an example, where we apply `CatToNumTransform <https://dl.acm.org/doi/10.1145/507533.507538>`_ to transform the categorical features into numerical features.