tests: update unit

Author: LongxingTan

.github/workflows/pypi_release.yml

@@ -4,8 +4,6 @@
 name: PyPi Release
 
 on:
-#  push:
-#    branches: [master]
   release:
     types: [created]
 

.github/workflows/test.yml

@@ -15,7 +15,7 @@ jobs:
     strategy:
       matrix:
         os: [ubuntu-latest, macOS-latest] # add windows-2019 when poetry allows installation with `-f` flag
-        python-version: [3.9, '3.11']
+        python-version: [3.9, '3.12']
         tf-version: [2.13.1, 2.15.1]
 
         exclude:
@@ -92,7 +92,7 @@ jobs:
       - name: Set up Python
         uses: actions/setup-python@v5
         with:
-          python-version: '3.10'
+          python-version: '3.12'
 
       - name: Create pip cache directory manually
         run: |

README.md

@@ -29,7 +29,7 @@
 [![Code Coverage][coverage-image]][coverage-url]
 [![Contributing][contributing-image]][contributing-url]
 
-**[Documentation](https://time-series-prediction.readthedocs.io)** | **[Tutorials](https://time-series-prediction.readthedocs.io/en/latest/tutorials.html)** | **[Release Notes](https://time-series-prediction.readthedocs.io/en/latest/CHANGELOG.html)** | **[中文](https://github.com/LongxingTan/Time-series-prediction/blob/master/README_CN.md)**
+**[Documentation](https://time-series-prediction.readthedocs.io)** | **[Tutorials](https://time-series-prediction.readthedocs.io/en/latest/tutorials.html)** | **[Release Notes](./CHANGELOG.md)** | **[中文](https://github.com/LongxingTan/Time-series-prediction/blob/master/README_CN.md)**
 
 **TFTS** (TensorFlow Time Series) is an easy-to-use time series package, supporting the classical and latest deep learning methods in TensorFlow or Keras.
 - Support sota models for time series tasks (prediction, classification, anomaly detection)

README_CN.md

@@ -29,7 +29,7 @@
 [![Code Coverage][coverage-image]][coverage-url]
 [![Contributing][contributing-image]][contributing-url]
 
-**[文档](https://time-series-prediction.readthedocs.io)** | **[教程](https://time-series-prediction.readthedocs.io/en/latest/tutorials.html)** | **[发布日志](https://time-series-prediction.readthedocs.io/en/latest/CHANGELOG.html)** | **[English](https://github.com/LongxingTan/Time-series-prediction/blob/master/README.md)**
+**[文档](https://time-series-prediction.readthedocs.io)** | **[教程](https://time-series-prediction.readthedocs.io/en/latest/tutorials.html)** | **[发布日志](./CHANGELOG.md)** | **[English](https://github.com/LongxingTan/Time-series-prediction/blob/master/README.md)**
 
 青山遮不住，毕竟东流去。江晚正愁余，山深闻鹧鸪。<br>
 **东流TFTS** (TensorFlow Time Series) 是一个高效易用的时间序列框架，基于TensorFlow/ Keras。

docs/source/models.rst

@@ -28,3 +28,16 @@ You can use below models with ``AutoModel``
 
     config = AutoConfig.for_model("seq2seq")
     model = AutoModel.from_config(config, predict_sequence_length=predict_sequence_length)
+
+
+Add a custom head for tfts model
+
+.. code-block:: python
+
+    config = AutoConfig.for_model("seq2seq")
+    model = AutoModel.from_config(config, predict_sequence_length=predict_sequence_length)
+    model.project = tf.keras.Sequential(
+        layers=[],
+        trainable=True,
+        name=None
+    )

docs/source/tricks.rst

@@ -45,3 +45,28 @@ There is no free launch, and it's impossible to forecast the future. So we shoul
     * add a hidden-sizes dense layer at last
     * encoder-decoder structure
     * encoder-forecasting structure
+
+
+.. code-block:: python
+
+    # use tfts auto-regressive generate multiple steps
+    from tfts.data import TimeSeriesSequence
+    # Generate predictions
+    last_sequence = data.tail(10)
+
+    # Function to add features after each prediction
+    def add_features(new_df, history_df):
+        # Add any features needed for the next prediction
+        # For example, you could add lag features, moving averages, etc.
+        return new_df
+
+    # Generate predictions
+    predictions = model.generate(
+        last_sequence,
+        generation_config={
+            'steps': 30,
+            'time_idx': 'time_idx',
+            'time_step': 1,
+            'add_features_func': add_features
+        }
+    )

tests/test_data/test_timeseries.py

@@ -1,6 +1,5 @@
 """Tests for TimeSeriesSequence class."""
 
-from datetime import datetime, timedelta
 import unittest
 
 import numpy as np

tests/test_examples/test_tfts_inputs.py

@@ -2,7 +2,6 @@
 import unittest
 
 import numpy as np
-import pandas as pd
 import tensorflow as tf
 
 from tfts import AutoConfig, AutoModel, KerasTrainer

tests/test_features/test_datetime_feature.py

@@ -1,6 +1,5 @@
 """Tests for datetime features."""
 
-from datetime import datetime, timedelta
 import unittest
 
 import numpy as np

tests/test_features/test_one_order_feature.py

@@ -1,6 +1,5 @@
 """Tests for one-order features."""
 
-from datetime import datetime, timedelta
 import unittest
 
 import numpy as np