STUMPY

https://stumpy.readthedocs.io/en/latest/
Admin
STUMPY is a powerful and scalable Python library for time series data mining that efficiently...
Tokens:2,144
Snippets:29
Trust Score:9.6
Update:5 months ago
Show doc for...
Tokens:
Raw
### Install STUMPY using pip

Source: https://stumpy.readthedocs.io/en/latest

Install STUMPY using pip. This command installs the package directly from PyPI.

```bash
python -m pip install stumpy

```

--------------------------------

### Install STUMPY using uv

Source: https://stumpy.readthedocs.io/en/latest

Add STUMPY to your project dependencies using the uv package manager.

```bash
uv add stumpy

```

--------------------------------

### Install STUMPY using Conda

Source: https://stumpy.readthedocs.io/en/latest

Install STUMPY from the conda-forge channel. This is a recommended installation method for conda users.

```bash
conda install -c conda-forge stumpy

```

--------------------------------

### Install STUMPY using Pixi

Source: https://stumpy.readthedocs.io/en/latest

Add STUMPY to your project dependencies using the Pixi package manager.

```bash
pixi add stumpy

```

--------------------------------

### Basic 1D Time Series Analysis with STUMPY

Source: https://stumpy.readthedocs.io/en/latest

Use this for typical 1-dimensional time series data analysis. Ensure you have NumPy and STUMPY installed.

```python
import stumpy
import numpy as np

if __name__ == "__main__":
    your_time_series = np.random.rand(10000)
    window_size = 50  # Approximately, how many data points might be found in a pattern

    matrix_profile = stumpy.stump(your_time_series, m=window_size)

```

--------------------------------

### Distributed 1D Time Series Analysis with STUMPED (Dask)

Source: https://stumpy.readthedocs.io/en/latest

Leverage Dask Distributed for parallel processing of 1-dimensional time series data. Requires Dask and STUMPY.

```python
import stumpy
import numpy as np
from dask.distributed import Client

if __name__ == "__main__":
    with Client() as dask_client:
        your_time_series = np.random.rand(10000)
        window_size = 50  # Approximately, how many data points might be found in a pattern

        matrix_profile = stumpy.stumped(dask_client, your_time_series, m=window_size)

```

--------------------------------

### Execute Tests with PyTest

Source: https://stumpy.readthedocs.io/en/latest

Run the project's tests using the provided shell script. This script is designed to be executed from the root of the project directory.

```bash
./test.sh

```

--------------------------------

### Distributed Multi-dimensional Time Series Analysis with MSTUMPED (Dask)

Source: https://stumpy.readthedocs.io/en/latest

Perform distributed analysis on multi-dimensional time series data using Dask and MSTUMPED. Requires Dask and STUMPY.

```python
import stumpy
import numpy as np
from dask.distributed import Client

if __name__ == "__main__":
    with Client() as dask_client:
        your_time_series = np.random.rand(3, 1000)   # Each row represents data from a different dimension while each column represents data from the same dimension
        window_size = 50  # Approximately, how many data points might be found in a pattern

        matrix_profile, matrix_profile_indices = stumpy.mstumped(dask_client, your_time_series, m=window_size)

```

--------------------------------

### Time Series Chains with Anchored Time Series Chains (ATSC)

Source: https://stumpy.readthedocs.io/en/latest

Extract anchored time series chains using ATSC. This function requires pre-computed matrix profiles and indices.

```python
import stumpy
import numpy as np

if __name__ == "__main__":
    your_time_series = np.random.rand(10000)
    window_size = 50  # Approximately, how many data points might be found in a pattern

    matrix_profile = stumpy.stump(your_time_series, m=window_size)

    left_matrix_profile_index = matrix_profile[:, 2]
    right_matrix_profile_index = matrix_profile[:, 3]
    idx = 10  # Subsequence index for which to retrieve the anchored time series chain for

    anchored_chain = stumpy.atsc(left_matrix_profile_index, right_matrix_profile_index, idx)

    all_chain_set, longest_unanchored_chain = stumpy.allc(left_matrix_profile_index, right_matrix_profile_index)

```

--------------------------------

### Semantic Segmentation with Fast Low-cost Unipotent Semantic Segmentation (FLUSS)

Source: https://stumpy.readthedocs.io/en/latest

Apply FLUSS for semantic segmentation of time series data. This function operates on a pre-computed matrix profile.

```python
import stumpy
import numpy as np

if __name__ == "__main__":
    your_time_series = np.random.rand(10000)
    window_size = 50  # Approximately, how many data points might be found in a pattern

    matrix_profile = stumpy.stump(your_time_series, m=window_size)

    subseq_len = 50
    correct_arc_curve, regime_locations = stumpy.fluss(matrix_profile[:, 1],
                                                    L=subseq_len,
                                                    n_regimes=2,
                                                    excl_factor=1
                                                    )

```

--------------------------------

### Multi-dimensional Time Series Analysis with MSTUMP

Source: https://stumpy.readthedocs.io/en/latest

Analyze multi-dimensional time series data using MSTUMP. Each row in the input array represents a different dimension.

```python
import stumpy
import numpy as np

if __name__ == "__main__":
    your_time_series = np.random.rand(3, 1000)  # Each row represents data from a different dimension while each column represents data from the same dimension
    window_size = 50  # Approximately, how many data points might be found in a pattern

    matrix_profile, matrix_profile_indices = stumpy.mstump(your_time_series, m=window_size)

```

--------------------------------

### GPU Accelerated 1D Time Series Analysis with GPU-STUMP

Source: https://stumpy.readthedocs.io/en/latest

Utilize GPU acceleration for 1-dimensional time series analysis. Requires Numba CUDA and STUMPY. Ensure you have compatible GPU hardware.

```python
import stumpy
import numpy as np
from numba import cuda

if __name__ == "__main__":
    your_time_series = np.random.rand(10000)
    window_size = 50  # Approximately, how many data points might be found in a pattern
    all_gpu_devices = [device.id for device in cuda.list_devices()]  # Get a list of all available GPU devices

    matrix_profile = stumpy.gpu_stump(your_time_series, m=window_size, device_id=all_gpu_devices)

```

--------------------------------

### BibTeX Citation for STUMPY

Source: https://stumpy.readthedocs.io/en/latest

Use this BibTeX entry to cite the STUMPY library in scientific publications. It includes author, title, journal, volume, number, pages, and year.

```bibtex
@article{law2019stumpy,
  author  = {Law, Sean M.},
  title   = {{STUMPY: A Powerful and Scalable Python Library for Time Series Data Mining}},
  journal = {{The Journal of Open Source Software}},
  volume  = {4},
  number  = {39},
  pages   = {1504},
  year    = {2019}
}

```
Tokens:
Raw