Keras 3.0

.gitignore

@@ -133,3 +133,9 @@ dmypy.json
 
 # Pyre type checker
 .pyre/
+
+# local history
+.lh/
+
+# data
+data/

docs/guides/augmentations.rst

@@ -0,0 +1,94 @@
+.. _augmentations-label:
+
+Augmentations
+============
+
+[Recommended] A more in-depth tutorial on augmentations for time series data `is available in our repo. <https://github.com/AlexanderVNikitin/tsgm/blob/main/tutorials/augmentations.ipynb>`_
+
+TSGM provides a wide variety of augmentation techniques beyond generative models.
+For the following demonstrations, we first need to generate a toy dataset:
+
+.. code-block:: python
+
+	import tsgm
+	X = tsgm.utils.gen_sine_dataset(100, 64, 2, max_value=20)
+
+Jittering
+------------
+In tsgm, Gaussian noise augmentation can be applied as follows:
+
+.. code-block:: python
+
+  aug_model = tsgm.models.augmentations.GaussianNoise()
+  samples = aug_model.generate(X=X, n_samples=10, variance=0.2)
+
+The idea behind Gaussian noise augmentation is that adding a small amount of jittering to time series probably will not change it significantly but will increase the amount of such noisy samples in our dataset.
+
+Shuffle Features
+------------
+Another approach to time series augmentation is simply shuffle the features. This approach is suitable only for particular multivariate time series, where they are invariant to all or particular permutations of features. For instance, it can be applied to time series where each feature represents same independent measurements from various sensors.
+
+.. code-block:: python
+
+  aug_model = tsgm.models.augmentations.Shuffle()
+  samples = aug_model.generate(X=X, n_samples=3)
+
+Slice and shuffle
+------------
+Slice and shuffle augmentation [3] cuts a time series into slices and shuffles those pieces. This augmentation can be performed for time series that exhibit some form of invariance over time. For instance, imagine a time series measured from wearable devices for several days. The good strategy for this case is to slice time series by days and, by shuffling those days, get additional samples. 
+
+.. code-block:: python
+
+  aug_model = tsgm.models.augmentations.SliceAndShuffle()
+  samples = aug_model.generate(X=X, n_samples=10, n_segments=3)
+
+Magnitude Warping
+------------
+Magnitude warping [3] changes the magnitude of each sample in a time series dataset by multiplication of the original time series with a cubic spline curve. This process scales the magnitude of time series, which can be beneficial in many cases, such as our synthetic example with sines n_knots number of knots at random magnitudes distributed as N(1, σ^2) where σ is set by a parameter sigma in function .generate.
+
+.. code-block:: python
+
+  aug_model = tsgm.models.augmentations.MagnitudeWarping()
+  samples = aug_model.generate(X=X, n_samples=10, sigma=1)  
+
+
+
+Window Warping
+------------
+In this technique [4], the selected windows in time series data are either speeding up or down. Then, the whole resulting time series is scaled back to the original size in order to keep the timesteps at the original length. See an example of such augmentation below:
+
+.. code-block:: python
+
+  aug_model = tsgm.models.augmentations.WindowWarping()
+  samples = aug_model.generate(X=X, n_samples=10, scales=(0.5,), window_ratio=0.5) 
+
+
+Dynamic Time Warping Barycentric Average (DTWBA)
+------------
+Dynamic Time Warping Barycentric Average (DTWBA)[2] is an augmentation method that is based on Dynamic Time Warping (DTW). DTW is a method of measuring similarity between time series. The idea is to "sync" those time series, as it is demonstrated in the following picture.
+
+DTWBA goes like this:
+
+  1. The algorithm picks one time series to initialize the DTW_BA result. 
+  2. This time series can either be given explicitly or can be chosen randomly from the dataset
+  3. For each of the N time series, the algorithm computes DTW distance and the path (the path is the mapping that minimizes the distance)
+  4. After computing all DTW distances, the algorithm updates the DTWBA result by doing the average with respect to all the paths found above
+  5. The algorithm repeats steps (2) and (3) until the DTWBA result converges
+
+.. code-block:: python
+
+  aug_model = tsgm.models.augmentations.DTWBarycentricAveraging()
+  initial_timeseries = random.sample(range(X.shape[0]), 10)
+  initial_timeseries = X[initial_timeseries]
+  samples = aug_model.generate(X=X, n_samples=10, initial_timeseries=initial_timeseries ) 
+
+
+References
+------------
+[1] H. Sakoe and S. Chiba, “Dynamic programming algorithm optimization for spoken word recognition”. IEEE Transactions on Acoustics, Speech, and Signal Processing, 26(1), 43-49 (1978).
+
+[2] F. Petitjean, A. Ketterlin & P. Gancarski. A global averaging method for dynamic time warping, with applications to clustering. Pattern Recognition, Elsevier, 2011, Vol. 44, Num. 3, pp. 678-693
+
+[3] Um TT, Pfister FM, Pichler D, Endo S, Lang M, Hirche S, Fietzek U, Kulic´ D (2017) Data augmentation of wearable sensor data for parkinson’s disease monitoring using convolutional neural networks. In: Proceedings of the 19th ACM international conference on multimodal interaction, pp. 216–220
+
+[4] Rashid, K.M. and Louis, J., 2019. Window-warping: a time series data augmentation of IMU data for construction equipment activity identification. In ISARC. Proceedings of the international symposium on automation and robotics in construction (Vol. 36, pp. 651-657). IAARC Publications.

docs/guides/introduction.rst

@@ -29,7 +29,7 @@ TSGM implements multiple augmentation approaches including window warping, shuff
 	aug_model = tsgm.models.augmentations.GaussianNoise(variance=0.2)
 	samples = aug_model.generate(X=X, n_samples=10)
 
-More examples are available in `the augmentation tutorial. <https://github.com/AlexanderVNikitin/tsgm/blob/main/tutorials/augmentations.ipynb>`_
+More examples are available in `the augmentation tutorial <https://github.com/AlexanderVNikitin/tsgm/blob/main/tutorials/augmentations.ipynb>`_ or in :ref:`augmentations-label`.
 
 Generators
 =============================
@@ -42,7 +42,7 @@ The training of data-driven simulators can be done via likelihood optimization,
 - `tsgm.models.cgan.ConditionalGAN` - conditional GAN model for labeled and temporally labeled time-series simulation,\\
 - `tsgm.models.cvae.BetaVAE` - beta-VAE model adapted for time-series simulation,\\
 - `tsgm.models.cvae.cBetaVAE` - conditional beta-VAE model for labeled and temporally labeled time-series simulation,\\
-- `tsgm.models.cvae.TimeGAN` - extended GAN-based model for time series generation.
+- `tsgm.models.timegan.TimeGAN` - extended GAN-based model for time series generation.
 
 A minimalistic example of synthetic data generation with VAEs:
 
@@ -103,12 +103,13 @@ Metrics
 =============================
 In `tsgm.metrics`, we implemented several metrics for evaluation of generated time series. Essentially, these metrics are subdivided into five types:
 
-- data similarity / distance,
-- predictive consistency,
-- fairness,
-- privacy,
-- downstream effectiveness,
-- visual similarity.
+- data similarity / distance: `tsgm.metrics.DistanceMetric`, `tsgm.metrics.MMDMetric`, `tsgm.metrics.DiscriminativeMetric`,
+- predictive consistency: `tsgm.metrics.ConsistencyMetric`,
+- fairness: `tsgm.metrics.DemographicParityMetric`, `tsgm.metrics.PredictiveParityMetric`
+- privacy: `tsgm.metrics.PrivacyMembershipInferenceMetric`,
+- diversity: `tsgm.metrics.EntropyMetric`, `tsgm.metrics.ShannonEntropyMetric`, `tsgm.metrics.PairwiseDistanceMetric`,
+- downstream effectiveness: `tsgm.metrics.DownstreamPerformanceMetric`,
+- qualitative analysis: `tsgm.utils.visualization`.
 
 See the following code for an example of using metrics:
 
@@ -151,5 +152,3 @@ If you find the *TSGM* useful, please consider citing our paper:
 	  journal={arXiv preprint arXiv:2305.11567},
 	  year={2023}
 	}
-
-

docs/modules/root.rst

@@ -11,7 +11,7 @@ Datasets
     :members:
     :undoc-members:
 
-.. automodule:: tsgm.utils.dataset
+.. automodule:: tsgm.utils.datasets
     :members:
     :undoc-members:
 
@@ -78,9 +78,9 @@ Zoo
     :undoc-members:
 
 
-Datasets
+Simulators
 --------------
-.. automodule:: tsgm.utils.datasets
+.. automodule:: tsgm.simulator
     :members:
     :undoc-members:
 

requirements/docs_requirements.txt

@@ -7,5 +7,5 @@ nbsphinx
 jupytext
 pydata-sphinx-theme
 sphinxcontrib-bibtex
-sphinx-autoapi
+sphinx-autoapi==3.0.0
 sphinx_rtd_theme

setup.py

@@ -1,8 +1,17 @@
+import os
 from setuptools import setup
 from setuptools import find_packages
 
 
+# Function to read version from __version__.py
+def get_version():
+    with open(os.path.join(os.path.dirname(__file__), 'tsgm/version.py')) as f:
+        exec(f.read())
+    return locals()['__version__']
+
+
 name = "tsgm"
+version = get_version()
 
 keywords = [
     "machine learning",
@@ -44,7 +53,7 @@ def read_file(filename: str) -> str:
 
 
 setup(name='tsgm',
-      version='0.0.4',
+      version=version,
       description='Time Series Generative Modelling Framework',
       author=author,
       author_email='',
@@ -70,8 +79,9 @@ def read_file(filename: str) -> str:
           "yfinance==0.2.28",
           "tqdm",
           "dtaidistance >= 2.3.10",
-          "tensorflow",
-          "tensorflow-probability",
+          "tensorflow < 2.16",
+          "tensorflow-probability < 0.24.0",
+          "statsmodels"
       ],
       package_data={'tsgm': ['README.md']},
       packages=find_packages())

tests/test_abc.py

@@ -1,10 +1,14 @@
 import pytest
+import os
+import sys
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
 import tsgm
-import tensorflow_probability as tfp
+from tsgm.backend import get_distributions
+
+distributions = get_distributions()
 
-import tensorflow as tf
 import numpy as np
-from tensorflow import keras
+import keras
 
 
 def test_abc_rejection_sampler_nn_simulator():
@@ -43,8 +47,8 @@ def test_abc_rejection_sampler_model_based_simulator():
     data = tsgm.dataset.DatasetProperties(N=100, D=2, T=100)
     simulator = tsgm.simulator.SineConstSimulator(data=data, max_scale=max_scale, max_const=20)
     priors = {
-        "max_scale": tfp.distributions.Uniform(9, 11),
-        "max_const": tfp.distributions.Uniform(19, 21)
+        "max_scale": distributions.Uniform(9, 11),
+        "max_const": distributions.Uniform(19, 21)
     }
     samples_ref = simulator.generate(10)
 

tests/test_augmentations.py

@@ -1,6 +1,8 @@
 import pytest
 import numpy as np
-
+import os
+import sys
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
 import tsgm
 
 @pytest.mark.parametrize("mock_aug", [

tests/test_cgan.py

@@ -1,14 +1,20 @@
 import pytest
+import os
+import sys
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
 import tsgm
 
-import tensorflow as tf
 try:
     import tensorflow_privacy as tf_privacy
     __tf_privacy_available = True
 except ModuleNotFoundError:
     __tf_privacy_available = False
 import numpy as np
-from tensorflow import keras
+import keras
+
+
+from tsgm.backend import get_backend
+
 
 
 def _gen_dataset(seq_len: int, feature_dim: int, batch_size: int):
@@ -17,8 +23,13 @@ def _gen_dataset(seq_len: int, feature_dim: int, batch_size: int):
     scaler = tsgm.utils.TSFeatureWiseScaler((-1, 1))
     X_train = scaler.fit_transform(data).astype(np.float32)
 
-    dataset = tf.data.Dataset.from_tensor_slices(X_train)
-    dataset = dataset.shuffle(buffer_size=1024).batch(batch_size)
+    backend = get_backend()
+    if os.environ.get("KERAS_BACKEND") == "tensorflow":
+        dataset = backend.data.Dataset.from_tensor_slices(X_train)
+        dataset = dataset.shuffle(buffer_size=1024).batch(batch_size)
+    elif os.environ.get("KERAS_BACKEND") == "torch":
+        dataset = backend.utils.data.TensorDataset(X_train)
+        dataset = backend.utils.data.DataLoader(dataset, batch_size=batch_size, shuffle=True)
     return dataset
 
 
@@ -29,8 +40,13 @@ def _gen_cond_dataset(seq_len: int, batch_size: int):
     X_train = scaler.fit_transform(X).astype(np.float32)
     y = keras.utils.to_categorical(y_i, 2).astype(np.float32)
 
-    dataset = tf.data.Dataset.from_tensor_slices((X_train, y))
-    dataset = dataset.shuffle(buffer_size=1024).batch(batch_size)
+    backend = get_backend()
+    if os.environ.get("KERAS_BACKEND") == "tensorflow":
+        dataset = backend.data.Dataset.from_tensor_slices(X_train)
+        dataset = dataset.shuffle(buffer_size=1024).batch(batch_size)
+    elif os.environ.get("KERAS_BACKEND") == "torch":
+        dataset = backend.utils.data.TensorDataset(X_train)
+        dataset = backend.utils.data.DataLoader(dataset, batch_size=batch_size, shuffle=True)
     return dataset, y
 
 
@@ -41,8 +57,13 @@ def _gen_t_cond_dataset(seq_len: int, batch_size: int):
     X_train = scaler.fit_transform(X).astype(np.float32)
     y = y.astype(np.float32)
 
-    dataset = tf.data.Dataset.from_tensor_slices((X_train, y))
-    dataset = dataset.shuffle(buffer_size=1024).batch(batch_size)
+    backend = get_backend()
+    if os.environ.get("KERAS_BACKEND") == "tensorflow":
+        dataset = backend.data.Dataset.from_tensor_slices(X_train)
+        dataset = dataset.shuffle(buffer_size=1024).batch(batch_size)
+    elif os.environ.get("KERAS_BACKEND") == "torch":
+        dataset = backend.utils.data.TensorDataset(X_train)
+        dataset = backend.utils.data.DataLoader(dataset, batch_size=batch_size, shuffle=True)
     return dataset, y
 
 
@@ -236,7 +257,6 @@ def test_dp_compiler():
         learning_rate=learning_rate
     )
 
-
     g_optimizer = tf_privacy.DPKerasAdamOptimizer(
         l2_norm_clip=l2_norm_clip,
         noise_multiplier=noise_multiplier,
@@ -259,6 +279,31 @@ def test_dp_compiler():
     assert generated_samples.shape == (10, 64, 1)    
 
 
-def test_temporal_cgan_multiple_features():
-    # TODO
-    pass
+def test_wavegan():
+    latent_dim = 2
+    output_dim = 1
+    feature_dim = 1
+    seq_len = 64
+    batch_size = 48
+
+    dataset = _gen_dataset(seq_len, feature_dim, batch_size)
+    architecture = tsgm.models.architectures.zoo["wavegan"](
+        seq_len=seq_len, feat_dim=feature_dim,
+        latent_dim=latent_dim, output_dim=output_dim)
+    discriminator, generator = architecture.discriminator, architecture.generator
+    gan = tsgm.models.cgan.GAN(
+        discriminator=discriminator, generator=generator, latent_dim=latent_dim, use_wgan=True
+    )
+    gan.compile(
+        d_optimizer=keras.optimizers.Adam(learning_rate=0.0003),
+        g_optimizer=keras.optimizers.Adam(learning_rate=0.0003),
+        loss_fn=keras.losses.BinaryCrossentropy(),
+    )
+
+    gan.fit(dataset, epochs=1)
+
+    assert gan.generator is not None
+    assert gan.discriminator is not None
+    # Check generation
+    generated_samples = gan.generate(10)
+    assert generated_samples.shape == (10, seq_len, 1)

tests/test_dataset.py

@@ -1,7 +1,11 @@
 import pytest
 
 import numpy as np
+import os
+import sys
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
 import tsgm
+import tsgm.backend
 
 
 def test_dataset():