Merge pull request #188 from WenjieDu/dev

Merge dev into main

.github/PULL_REQUEST_TEMPLATE.md

@@ -1,3 +1,5 @@
+<!-- 🚨Please create your PR to merge code from your branch to `dev` branch here, rather than `main`. -->
+
 # What does this PR do?
 
 <!--

.github/workflows/greetings.yml

@@ -1,4 +1,4 @@
-name: Autoreply to Issues Opened for PyPOTS
+name: Auto reply to new issues and PR
 
 on:
   issues:
@@ -28,7 +28,6 @@ jobs:
 
           Best,
           Wenjie
-
         pr-message: |
           Hi there 👋,
 

.gitignore

@@ -0,0 +1,20 @@
+# ignore special files or folds
+*~
+.idea
+.DS_Store
+
+# ignore all building results
+dist
+build
+docs/_build
+*.egg-info
+
+# ignore all testing/running results
+.run
+.coverage
+.pytest_cache
+*__pycache__*
+*testing_results*
+
+# ignore specific kinds of files like all PDFs
+*.pdf

.readthedocs.yaml

@@ -23,7 +23,7 @@ build:
          - pip install torch==1.13.1 -f https://download.pytorch.org/whl/cpu
          - pip install torch-geometric torch-scatter torch-sparse -f "https://data.pyg.org/whl/torch-1.13.1+cpu.html"
          - pip install pypots
-         - pip install sphinx==6.2.1 docutils==0.19 sphinxcontrib-bibtex==2.1.4 sphinxcontrib-gtagjs sphinx-autodoc-typehints furo
+         - pip install sphinx==6.2.1 docutils==0.19 sphinxcontrib-bibtex==2.1.4 sphinxcontrib-gtagjs sphinx-autodoc-typehints furo==2023.07.26
 
       post_install:
          - pip install docutils==0.20

MANIFEST.in

@@ -1,2 +1,3 @@
 prune tests
 prune pypots/*/template
+prune pypots/*/README.md

README.md

@@ -1,5 +1,5 @@
 <a href="https://github.com/WenjieDu/PyPOTS">
-    <img src="https://raw.githubusercontent.com/PyPOTS/pypots.github.io/main/static/figs/pypots_logos/PyPOTS_logo_FFBG.svg?sanitize=true" width="200" align="right">
+    <img src="https://pypots.com/figs/pypots_logos/PyPOTS_logo_FFBG.svg?sanitize=true" width="200" align="right">
 </a>
 
 ## <p align="center">Welcome to PyPOTS</p>
@@ -9,12 +9,14 @@
     <a href="https://docs.pypots.com/en/latest/install.html#reasons-of-version-limitations-on-dependencies">
        <img alt="Python version" src="https://img.shields.io/badge/Python-v3.7--3.10-E97040?logo=python&logoColor=white">
     </a>
-    <img alt="powered by Pytorch" src="https://img.shields.io/badge/PyTorch-❤️-F8C6B5?logo=pytorch&logoColor=white">
-    <a href="https://pypi.org/project/PyPOTS">
-        <img alt="the latest release version" src="https://img.shields.io/github/v/release/wenjiedu/pypots?color=EE781F&include_prereleases&label=Release">
+    <a href="https://github.com/WenjieDu/PyPOTS">
+        <img alt="powered by Pytorch" src="https://img.shields.io/badge/PyTorch-❤️-F8C6B5?logo=pytorch&logoColor=white">
+    </a>
+    <a href="https://github.com/WenjieDu/PyPOTS/releases">
+        <img alt="the latest release version" src="https://img.shields.io/github/v/release/wenjiedu/pypots?color=EE781F&include_prereleases&label=Release&logo=github&logoColor=white">
     </a>
     <a href="https://github.com/WenjieDu/PyPOTS/blob/main/LICENSE">
-        <img alt="GPL-v3 license" src="https://img.shields.io/badge/License-GPL--v3-E9BB41">
+        <img alt="GPL-v3 license" src="https://img.shields.io/badge/License-GPL--v3-E9BB41?logo=opensourceinitiative&logoColor=white">
     </a>
     <a href="https://github.com/WenjieDu/PyPOTS/blob/main/README.md#-community">
         <img alt="Community" src="https://img.shields.io/badge/join_us-community!-C8A062">
@@ -41,10 +43,10 @@
         <img alt="arXiv DOI" src="https://img.shields.io/badge/DOI-10.48550/arXiv.2305.18811-F8F7F0">
     </a>
     <a href="https://anaconda.org/conda-forge/pypots">
-        <img alt="Conda downloads" src="https://img.shields.io/conda/dn/conda-forge/pypots?label=Conda%20Downloads&color=AED0ED&logo=anaconda&logoColor=white">
+        <img alt="Conda downloads" src="https://img.shields.io/endpoint?url=https://pypots.com/figs/downloads_badges/conda_pypots_downloads.json">
     </a>
     <a href="https://pepy.tech/project/pypots">
-        <img alt="PyPI downloads" src="https://static.pepy.tech/personalized-badge/pypots?period=total&units=international_system&left_color=grey&right_color=blue&left_text=PyPI%20Downloads">
+        <img alt="PyPI downloads" src="https://img.shields.io/endpoint?url=https://pypots.com/figs/downloads_badges/pypi_pypots_downloads.json">
     </a>
 
 </p>
@@ -66,7 +68,7 @@ have unified APIs together with detailed documentation and interactive examples
 if it helps with your research. This really means a lot to our open-source research. Thank you!
 
 <a href="https://github.com/WenjieDu/TSDB">
-    <img src="https://raw.githubusercontent.com/PyPOTS/pypots.github.io/main/static/figs/pypots_logos/TSDB_logo_FFBG.svg?sanitize=true" align="left" width="160" alt="TSDB logo"/>
+    <img src="https://pypots.com/figs/pypots_logos/TSDB_logo_FFBG.svg?sanitize=true" align="left" width="160" alt="TSDB logo"/>
 </a>
 
 To make various open-source time-series datasets readily available to our users,
@@ -108,7 +110,7 @@ Alternatively, you can install from the latest source code with the latest featu
 
 ## ❖ Usage
 <a href="https://github.com/WenjieDu/BrewPOTS">
-    <img src="https://raw.githubusercontent.com/PyPOTS/pypots.github.io/main/static/figs/pypots_logos/BrewPOTS_logo_FFBG.svg?sanitize=true" align="left" width="160" alt="BrewPOTS logo"/>
+    <img src="https://pypots.com/figs/pypots_logos/BrewPOTS_logo_FFBG.svg?sanitize=true" align="left" width="160" alt="BrewPOTS logo"/>
 </a>
 
 PyPOTS tutorials have been released. Considering the future workload, I separate the tutorials into a single repo,
@@ -269,3 +271,4 @@ PyPOTS community is open, transparent, and surely friendly. Let's work together
     <img alt="PyPOTS visits" align="left" src="https://hits.seeyoufarm.com/api/count/incr/badge.svg?url=https%3A%2F%2Fgithub.com%2FPyPOTS%2FPyPOTS&count_bg=%23009A0A&title_bg=%23555555&icon=&icon_color=%23E7E7E7&title=Visits%20since%20May%202022&edge_flat=false">
 </a>
 </details>
+<br>

docs/examples.rst

@@ -1,7 +1,7 @@
 Quick-start Examples
 ====================
 
-.. image:: https://raw.githubusercontent.com/PyPOTS/pypots.github.io/main/static/figs/pypots_logos/BrewPOTS_logo_FFBG.svg?sanitize=true
+.. image:: https://pypots.com/figs/pypots_logos/BrewPOTS_logo_FFBG.svg?sanitize=true
    :width: 160
    :alt: BrewPOTS logo
    :align: right

docs/faq.rst

@@ -38,7 +38,7 @@ and
 
 **NOTE**: The maintainer role is not permanent. The role is called "maintainer" because it actively maintains the project.
 You can take a leave of absence from the role with notice at any time.
-But if you're inactive for a long time (more than three months. With reasons, a longer period is allowed for sure), your role will be deactivated.
+But if you're inactive for a long time (more than three months. With reasons, a longer period is allowed for sure), your role may be deactivated.
 
 Becoming a Lead
 """""""""""""""
@@ -52,7 +52,8 @@ The lead is a permanent role unless your research is no longer related to the fi
 you no longer want to get involved with affairs at PyPOTS.
 
 If you believe you want to do this, you can drop an email with anything you want to tell and your CV attachment to
-`[email protected] <mailto:[email protected]>`_. We will schedule a meeting for you and all other leads at PyPOTS for further discussion.
+`[email protected] <mailto:[email protected]>`_. We will schedule a meeting for you and all other members at PyPOTS for further discussion.
+This is absolutely not a so-called interview, please don't take it formal and we just would like to listen to your thoughts about the field of POTS ;-)
 
 
 Our Development Principles
@@ -64,7 +65,7 @@ Our Development Principles
    We develop PyPOTS and we should try the best to use it in any scenarios related to POTS data.
    Only in this way, we can figure out how it tastes like, if it is a good toolset for users, and what other features and models should be included into PyPOTS;
 3. `No silver bullet <https://en.wikipedia.org/wiki/No_Silver_Bullet>`_ and `No free launch <https://en.wikipedia.org/wiki/No_free_lunch_theorem>`_.
-   There is no one solution to all problems in the Universe. In PyPOTS, we keep things modular, so one can easily try and replace parts of the pipeline
+   There is no one solution to all problems in the universe. In PyPOTS, we keep things modular, so one can easily try and replace parts of the pipeline
    in search for the optimal combination for the particular task;
 4. Keep things easy to use and familiar. We try to keep PyPOTS intuitive without compromising flexibility and without forcing users to learn a completely new technology.
    We do this by keeping the toolkit close to APIs in scikit-learn and pytorch that people know and love;

docs/index.rst

@@ -4,25 +4,27 @@
 
 Welcome to PyPOTS docs!
 ===============================
-.. image:: https://raw.githubusercontent.com/PyPOTS/pypots.github.io/main/static/figs/pypots_logos/PyPOTS_logo_FFBG.svg?sanitize=true
+.. image:: https://pypots.com/figs/pypots_logos/PyPOTS_logo_FFBG.svg?sanitize=true
    :height: 168
    :align: right
    :target: https://github.com/WenjieDu/PyPOTS
    :alt: PyPOTS logo
 
-.. centered:: A Python Toolbox for Data Mining on Partially-Observed Time Series
+**A Python Toolbox for Data Mining on Partially-Observed Time Series**
 
 .. image:: https://img.shields.io/badge/Python-v3.7--3.10-E97040?logo=python&logoColor=white
    :alt: Python version
+   :target: https://docs.pypots.com/en/latest/install.html#reasons-of-version-limitations-on-dependencies
 
 .. image:: https://img.shields.io/badge/PyTorch-❤️-F8C6B5?logo=pytorch&logoColor=white
    :alt: powered by Pytorch
+   :target: https://github.com/WenjieDu/PyPOTS
 
-.. image:: https://img.shields.io/github/v/release/wenjiedu/pypots?color=EE781F&include_prereleases&label=Release
+.. image:: https://img.shields.io/github/v/release/wenjiedu/pypots?color=EE781F&include_prereleases&label=Release&logo=github&logoColor=white
    :alt: the latest release version
-   :target: https://pypi.org/project/pypots
+   :target: https://github.com/WenjieDu/PyPOTS/releases
 
-.. image:: https://img.shields.io/badge/License-GPL--v3-E9BB41
+.. image:: https://img.shields.io/badge/License-GPL--v3-E9BB41?logo=opensourceinitiative&logoColor=white
    :alt: GPL-v3 license
    :target: https://github.com/WenjieDu/PyPOTS/blob/main/LICENSE
 
@@ -58,11 +60,11 @@ Welcome to PyPOTS docs!
    :alt: arXiv DOI
    :target: https://arxiv.org/abs/2305.18811
 
-.. image:: https://img.shields.io/conda/dn/conda-forge/pypots?label=Conda%20Downloads&color=AED0ED&logo=anaconda&logoColor=white
+.. image:: https://img.shields.io/endpoint?url=https://pypots.com/figs/downloads_badges/conda_pypots_downloads.json
    :alt: Conda downloads
    :target: https://anaconda.org/conda-forge/pypots
 
-.. image:: https://static.pepy.tech/personalized-badge/pypots?period=total&units=international_system&left_color=grey&right_color=blue&left_text=PyPI%20Downloads
+.. image:: https://img.shields.io/endpoint?url=https://pypots.com/figs/downloads_badges/pypi_pypots_downloads.json
    :alt: PyPI downloads
    :target: https://pepy.tech/project/pypots
 
@@ -80,7 +82,7 @@ Welcome to PyPOTS docs!
 **Please** properly `cite PyPOTS <https://docs.pypots.com/en/latest/milestones.html#citing-pypots>`_ in your publications
 if it helps with your research. This really means a lot to our open-source research. Thank you!
 
-.. image:: https://raw.githubusercontent.com/PyPOTS/pypots.github.io/main/static/figs/pypots_logos/TSDB_logo_FFBG.svg?sanitize=true
+.. image:: https://pypots.com/figs/pypots_logos/TSDB_logo_FFBG.svg?sanitize=true
    :width: 170
    :alt: TSDB
    :align: left
@@ -108,7 +110,7 @@ Refer to the page `Installation <install.html>`_ to see different ways of instal
 
 ❖ Usage
 ^^^^^^^^
-.. image:: https://raw.githubusercontent.com/PyPOTS/pypots.github.io/main/static/figs/pypots_logos/BrewPOTS_logo_FFBG.svg?sanitize=true
+.. image:: https://pypots.com/figs/pypots_logos/BrewPOTS_logo_FFBG.svg?sanitize=true
    :width: 160
    :alt: BrewPOTS logo
    :align: left

docs/pypots.utils.rst

@@ -1,10 +1,10 @@
 pypots.utils package
 ====================
 
-pypots.utils.files module
+pypots.utils.file module
 -------------------------
 
-.. automodule:: pypots.utils.files
+.. automodule:: pypots.utils.file
    :members:
    :undoc-members:
    :show-inheritance:

pypots/data/__init__.py

@@ -22,6 +22,7 @@
     pickle_load,
     pickle_dump,
 )
+from .saving import save_dict_into_h5
 
 __all__ = [
     # datasets
@@ -39,4 +40,6 @@
     "mcar",
     "pickle_load",
     "pickle_dump",
+    # saving
+    "save_dict_into_h5",
 ]

pypots/data/saving.py

@@ -0,0 +1,43 @@
+"""
+Data saving utilities.
+"""
+
+# Created by Wenjie Du <[email protected]>
+# License: GLP-v3
+
+
+import os
+
+import h5py
+
+from pypots.utils.file import create_dir_if_not_exist
+from pypots.utils.logging import logger
+
+
+def save_dict_into_h5(data_dict: dict, saving_dir: str) -> None:
+    """Save the given data (in a dictionary) into the given h5 file.
+
+    Parameters
+    ----------
+    data_dict : dict,
+        The data to be saved, should be a Python dictionary.
+
+    saving_dir : str,
+        The h5 file to save the data.
+
+    """
+
+    def save_set(handle, name, data):
+        if isinstance(data, dict):
+            single_set_handle = handle.create_group(name)
+            for key, value in data.items():
+                save_set(single_set_handle, key, value)
+        else:
+            handle.create_dataset(name, data=data)
+
+    create_dir_if_not_exist(saving_dir)
+    saving_path = os.path.join(saving_dir, "datasets.h5")
+    with h5py.File(saving_path, "w") as hf:
+        for k, v in data_dict.items():
+            save_set(hf, k, v)
+    logger.info(f"Successfully saved the given data into {saving_path}.")

pypots/data/utils.py

@@ -192,3 +192,46 @@ def cal_delta_for_single_sample(mask: np.ndarray) -> np.ndarray:
             delta_collector.append(delta)
         delta = np.asarray(delta_collector)
     return delta
+
+
+def sliding_window(time_series, n_steps, sliding_len=None):
+    """Generate time series samples with sliding window method, truncating windows from time-series data
+    with a given sequence length.
+
+    Given a time series of shape [seq_len, n_features] (seq_len is the total sequence length of the time series), this
+    sliding_window function will generate time-series samples from this given time series with sliding window method.
+    The number of generated samples is seq_len//sliding_len. And the final returned numpy ndarray has a shape
+    [seq_len//sliding_len, n_steps, n_features].
+
+    Parameters
+    ----------
+    time_series : np.ndarray,
+        time series data, len(shape)=2, [total_length, feature_num]
+
+    n_steps : int,
+        The number of time steps in the generated data samples.
+
+    sliding_len : int, default = None,
+        The size of the sliding window. It will be set as the same with n_steps if None.
+
+    Returns
+    -------
+    samples : np.ndarray,
+        The generated time-series data samples of shape [seq_len//sliding_len, n_steps, n_features].
+
+    """
+    sliding_len = n_steps if sliding_len is None else sliding_len
+    total_len = time_series.shape[0]
+    start_indices = np.asarray(range(total_len // sliding_len)) * sliding_len
+
+    # remove the last one if left length is not enough
+    if total_len - start_indices[-1] * sliding_len < n_steps:
+        start_indices = start_indices[:-1]
+
+    sample_collector = []
+    for idx in start_indices:
+        sample_collector.append(time_series[idx : idx + n_steps])
+
+    samples = np.asarray(sample_collector).astype("float32")
+
+    return samples

pypots/imputation/saits/model.py

@@ -156,7 +156,7 @@ def forward(
         if (training and self.diagonal_attention_mask) or (
             (not training) and diagonal_attention_mask
         ):
-            diagonal_attention_mask = torch.eye(self.n_steps).to(X.device)
+            diagonal_attention_mask = (1 - torch.eye(self.n_steps)).to(X.device)
             # then broadcast on the batch axis
             diagonal_attention_mask = diagonal_attention_mask.unsqueeze(0)
         else:

pypots/modules/self_attention.py

@@ -42,7 +42,7 @@ def forward(
 
         # apply masking on the attention map, this is optional
         if attn_mask is not None:
-            attn = attn.masked_fill(attn_mask == 1, -1e9)
+            attn = attn.masked_fill(attn_mask == 0, -1e9)
 
         # compute attention score [0, 1], then apply dropout
         attn = self.dropout(F.softmax(attn, dim=-1))
@@ -207,10 +207,10 @@ def forward(
         dec_enc_attn_mask: Optional[torch.Tensor] = None,
     ) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
         dec_output, dec_slf_attn = self.slf_attn(
-            dec_input, dec_input, dec_input, mask=slf_attn_mask
+            dec_input, dec_input, dec_input, attn_mask=slf_attn_mask
         )
         dec_output, dec_enc_attn = self.enc_attn(
-            dec_output, enc_output, enc_output, mask=dec_enc_attn_mask
+            dec_output, enc_output, enc_output, attn_mask=dec_enc_attn_mask
         )
         dec_output = self.pos_ffn(dec_output)
         return dec_output, dec_slf_attn, dec_enc_attn
@@ -288,7 +288,7 @@ def __init__(
         self.embedding = nn.Linear(n_features, d_model)
         self.dropout = nn.Dropout(dropout)
         self.position_enc = PositionalEncoding(d_model, n_position=n_steps)
-        self.dec_layer_stack = nn.ModuleList(
+        self.layer_stack = nn.ModuleList(
             [
                 DecoderLayer(
                     d_model,