ts-quant 0.1.0__tar.gz

ts_quant-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024-2026 TS-Quant Team
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

ts_quant-0.1.0/PKG-INFO

@@ -0,0 +1,192 @@
+Metadata-Version: 2.4
+Name: ts-quant
+Version: 0.1.0
+Summary: GPU-accelerated quantitative time-series feature extraction for financial data
+Author: AnalisaSaham
+License: MIT
+Keywords: time-series,feature-extraction,gpu,quantitative-finance,catboost
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Financial and Insurance Industry
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Topic :: Office/Business :: Financial :: Investment
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: torch>=2.0.0
+Requires-Dist: numpy>=1.24.0
+Requires-Dist: pandas>=1.5.0
+Requires-Dist: tqdm>=4.65.0
+Provides-Extra: signatures
+Requires-Dist: signatory>=1.2.6; extra == "signatures"
+Provides-Extra: test
+Requires-Dist: pytest>=7.0; extra == "test"
+Requires-Dist: pycatch22; extra == "test"
+Requires-Dist: PyWavelets; extra == "test"
+Dynamic: license-file
+
+# ts-quant 🚀
+
+**GPU-Accelerated Quantitative Feature Extraction for Financial Time Series**
+
+[![PyPI](https://img.shields.io/pypi/v/ts-quant)](https://pypi.org/project/ts-quant/)
+[![Python 3.9+](https://img.shields.io/badge/python-3.9+-blue.svg)](https://www.python.org/downloads/)
+[![PyTorch](https://img.shields.io/badge/PyTorch-2.0+-ee4c2c.svg)](https://pytorch.org/)
+[![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](LICENSE)
+
+Extract **2000+ quantitative features** from OHLCV stock data using 5 powerful engines — all running on GPU via PyTorch.
+
+## 🎯 Why ts-quant?
+
+| Feature | tsfresh (CPU) | ts-quant (GPU) |
+|---|---|---|
+| Speed | ~60s per stock | ~0.3s per stock |
+| Memory | OOM on large data | VRAM-managed, crash-proof |
+| Features | 794 (statistical only) | 2000+ (statistical + wavelets + rocket + signatures) |
+| GPU | ❌ | ✅ PyTorch/CUDA |
+
+## 📦 Installation
+
+```bash
+pip install ts-quant
+```
+
+**Requirements:** Python 3.9+, PyTorch 2.0+ (with CUDA for GPU acceleration)
+
+## 🚀 Quick Start
+
+```python
+import pandas as pd
+from ts_quant import generate_features
+
+# Load your OHLCV data
+df = pd.read_csv('stocks.csv')
+# Expected columns: symbol, date, close, volume, open, high, low
+
+# Extract features (GPU)
+df_features = generate_features(
+    df,
+    device='cuda',              # 'cpu' for CPU-only
+    tsfresh_mode='comprehensive',
+    window_sizes=[20],
+)
+
+print(f"New features: {df_features.shape[1] - df.shape[1]}")
+```
+
+## 🔧 5 Feature Engines
+
+### Engine A: MultiRocket 🎯
+Random convolutional kernels with multi-scale dilations.
+```python
+from ts_quant import RocketEngine
+
+engine = RocketEngine(n_kernels=250)
+features, names = engine.extract(x)  # x: [B, T]
+# → 1000 features (250 kernels × 4 pooling ops)
+```
+
+### Engine B: Catch22 📊
+22 canonical time-series features (ACF, DFA, entropy, etc.).
+```python
+from ts_quant import Catch22Engine
+
+engine = Catch22Engine()
+features, names = engine.extract(x)
+# → 22 features per window
+```
+
+### Engine C: Path Signatures ✍️
+Ordered interaction features via iterated integrals.
+```python
+from ts_quant import SignaturesEngine
+
+engine = SignaturesEngine(depth=3, channels=['close', 'volume'])
+features, names = engine.extract(x_multivariate)  # [B, T, d]
+# → d + d² + d³ features
+```
+
+### Engine D: Wavelets 🌊
+Discrete wavelet transform (Haar, db4, db2, sym4).
+```python
+from ts_quant import WaveletsEngine
+
+engine = WaveletsEngine(wavelet_types=['haar', 'db4'])
+features, names = engine.extract(x)
+# → 62 features (energy, entropy, mean, std, max per level)
+```
+
+### Engine E: Tsfresh Complete 📈
+63 mathematical functions, 361 features in comprehensive mode.
+```python
+from ts_quant import TsfreshEngine
+
+engine = TsfreshEngine(mode='comprehensive')
+features, names = engine.extract(x)
+# → 361 features (statistics, ACF, FFT, entropy, trend, ...)
+```
+
+## ⚡ VRAM Management
+
+ts-quant automatically manages GPU memory to prevent OOM crashes:
+
+```python
+# Works on any GPU size (8GB - 80GB)
+df_features = generate_features(
+    df,
+    device='cuda',
+    max_vram_gb=8,  # Optional: manual VRAM limit
+)
+```
+
+## 🔬 Feature Selection
+
+Built-in redundancy removal:
+
+```python
+from ts_quant import auto_select_features
+
+selected, names = auto_select_features(
+    features_tensor,
+    feature_names,
+    correlation_threshold=0.95,  # Remove highly correlated
+)
+```
+
+## 📁 Input Format
+
+Your DataFrame should have this structure:
+
+| symbol | date | open | high | low | close | volume |
+|--------|------|------|------|-----|-------|--------|
+| BBCA | 2024-01-02 | 9500 | 9600 | 9400 | 9550 | 1000000 |
+| BBRI | 2024-01-02 | 5200 | 5300 | 5100 | 5250 | 2000000 |
+
+## 🏗️ Architecture
+
+```
+ts-quant/
+├── ts_quant/
+│   ├── api.py              # Orchestrator — generate_features()
+│   ├── core/
+│   │   ├── memory_manager.py   # Dynamic VRAM management
+│   │   ├── tensor_utils.py     # DataFrame ↔ Tensor conversion
+│   │   └── windowing.py        # GPU sliding windows
+│   ├── engines/
+│   │   ├── rocket.py           # Engine A: MultiRocket
+│   │   ├── catch22.py          # Engine B: Catch22
+│   │   ├── signatures.py       # Engine C: Path Signatures
+│   │   ├── wavelets.py         # Engine D: Wavelets
+│   │   └── tsfresh_core.py     # Engine E: Tsfresh Complete
+│   └── utils/
+│       ├── config.py           # Default configurations
+│       ├── validation.py       # Input validation
+│       └── feature_selection.py # Redundancy removal
+└── tests/
+```
+
+## 📄 License
+
+MIT License

ts_quant-0.1.0/README.md

@@ -0,0 +1,163 @@
+# ts-quant 🚀
+
+**GPU-Accelerated Quantitative Feature Extraction for Financial Time Series**
+
+[![PyPI](https://img.shields.io/pypi/v/ts-quant)](https://pypi.org/project/ts-quant/)
+[![Python 3.9+](https://img.shields.io/badge/python-3.9+-blue.svg)](https://www.python.org/downloads/)
+[![PyTorch](https://img.shields.io/badge/PyTorch-2.0+-ee4c2c.svg)](https://pytorch.org/)
+[![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](LICENSE)
+
+Extract **2000+ quantitative features** from OHLCV stock data using 5 powerful engines — all running on GPU via PyTorch.
+
+## 🎯 Why ts-quant?
+
+| Feature | tsfresh (CPU) | ts-quant (GPU) |
+|---|---|---|
+| Speed | ~60s per stock | ~0.3s per stock |
+| Memory | OOM on large data | VRAM-managed, crash-proof |
+| Features | 794 (statistical only) | 2000+ (statistical + wavelets + rocket + signatures) |
+| GPU | ❌ | ✅ PyTorch/CUDA |
+
+## 📦 Installation
+
+```bash
+pip install ts-quant
+```
+
+**Requirements:** Python 3.9+, PyTorch 2.0+ (with CUDA for GPU acceleration)
+
+## 🚀 Quick Start
+
+```python
+import pandas as pd
+from ts_quant import generate_features
+
+# Load your OHLCV data
+df = pd.read_csv('stocks.csv')
+# Expected columns: symbol, date, close, volume, open, high, low
+
+# Extract features (GPU)
+df_features = generate_features(
+    df,
+    device='cuda',              # 'cpu' for CPU-only
+    tsfresh_mode='comprehensive',
+    window_sizes=[20],
+)
+
+print(f"New features: {df_features.shape[1] - df.shape[1]}")
+```
+
+## 🔧 5 Feature Engines
+
+### Engine A: MultiRocket 🎯
+Random convolutional kernels with multi-scale dilations.
+```python
+from ts_quant import RocketEngine
+
+engine = RocketEngine(n_kernels=250)
+features, names = engine.extract(x)  # x: [B, T]
+# → 1000 features (250 kernels × 4 pooling ops)
+```
+
+### Engine B: Catch22 📊
+22 canonical time-series features (ACF, DFA, entropy, etc.).
+```python
+from ts_quant import Catch22Engine
+
+engine = Catch22Engine()
+features, names = engine.extract(x)
+# → 22 features per window
+```
+
+### Engine C: Path Signatures ✍️
+Ordered interaction features via iterated integrals.
+```python
+from ts_quant import SignaturesEngine
+
+engine = SignaturesEngine(depth=3, channels=['close', 'volume'])
+features, names = engine.extract(x_multivariate)  # [B, T, d]
+# → d + d² + d³ features
+```
+
+### Engine D: Wavelets 🌊
+Discrete wavelet transform (Haar, db4, db2, sym4).
+```python
+from ts_quant import WaveletsEngine
+
+engine = WaveletsEngine(wavelet_types=['haar', 'db4'])
+features, names = engine.extract(x)
+# → 62 features (energy, entropy, mean, std, max per level)
+```
+
+### Engine E: Tsfresh Complete 📈
+63 mathematical functions, 361 features in comprehensive mode.
+```python
+from ts_quant import TsfreshEngine
+
+engine = TsfreshEngine(mode='comprehensive')
+features, names = engine.extract(x)
+# → 361 features (statistics, ACF, FFT, entropy, trend, ...)
+```
+
+## ⚡ VRAM Management
+
+ts-quant automatically manages GPU memory to prevent OOM crashes:
+
+```python
+# Works on any GPU size (8GB - 80GB)
+df_features = generate_features(
+    df,
+    device='cuda',
+    max_vram_gb=8,  # Optional: manual VRAM limit
+)
+```
+
+## 🔬 Feature Selection
+
+Built-in redundancy removal:
+
+```python
+from ts_quant import auto_select_features
+
+selected, names = auto_select_features(
+    features_tensor,
+    feature_names,
+    correlation_threshold=0.95,  # Remove highly correlated
+)
+```
+
+## 📁 Input Format
+
+Your DataFrame should have this structure:
+
+| symbol | date | open | high | low | close | volume |
+|--------|------|------|------|-----|-------|--------|
+| BBCA | 2024-01-02 | 9500 | 9600 | 9400 | 9550 | 1000000 |
+| BBRI | 2024-01-02 | 5200 | 5300 | 5100 | 5250 | 2000000 |
+
+## 🏗️ Architecture
+
+```
+ts-quant/
+├── ts_quant/
+│   ├── api.py              # Orchestrator — generate_features()
+│   ├── core/
+│   │   ├── memory_manager.py   # Dynamic VRAM management
+│   │   ├── tensor_utils.py     # DataFrame ↔ Tensor conversion
+│   │   └── windowing.py        # GPU sliding windows
+│   ├── engines/
+│   │   ├── rocket.py           # Engine A: MultiRocket
+│   │   ├── catch22.py          # Engine B: Catch22
+│   │   ├── signatures.py       # Engine C: Path Signatures
+│   │   ├── wavelets.py         # Engine D: Wavelets
+│   │   └── tsfresh_core.py     # Engine E: Tsfresh Complete
+│   └── utils/
+│       ├── config.py           # Default configurations
+│       ├── validation.py       # Input validation
+│       └── feature_selection.py # Redundancy removal
+└── tests/
+```
+
+## 📄 License
+
+MIT License

ts_quant-0.1.0/pyproject.toml

@@ -0,0 +1,43 @@
+[build-system]
+requires = ["setuptools>=68.0", "wheel"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "ts-quant"
+version = "0.1.0"
+description = "GPU-accelerated quantitative time-series feature extraction for financial data"
+readme = "README.md"
+license = {text = "MIT"}
+requires-python = ">=3.8"
+authors = [
+    {name = "AnalisaSaham"},
+]
+keywords = ["time-series", "feature-extraction", "gpu", "quantitative-finance", "catboost"]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Financial and Insurance Industry",
+    "Intended Audience :: Science/Research",
+    "License :: OSI Approved :: MIT License",
+    "Programming Language :: Python :: 3",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence",
+    "Topic :: Office/Business :: Financial :: Investment",
+]
+
+dependencies = [
+    "torch>=2.0.0",
+    "numpy>=1.24.0",
+    "pandas>=1.5.0",
+    "tqdm>=4.65.0",
+]
+
+[project.optional-dependencies]
+signatures = ["signatory>=1.2.6"]
+test = [
+    "pytest>=7.0",
+    "pycatch22",
+    "PyWavelets",
+]
+
+[tool.setuptools.packages.find]
+where = ["."]
+include = ["ts_quant*"]

ts_quant-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+