Getting started¶

Setting up¶

With pip¶

pymdea is available on PyPI and can be installed with:

pip install pymdea

With uv¶

pymdea can also be installed from PyPI with uv:

uv add pymdea

With uv from source¶

Install uv¶

Install the uv python project manager, following the instructions on uv's official installation guide.

Clone pymdea¶

Using a command line terminal, clone the pymdea repository to a location on your file system:

git clone https://github.com/garland-culbreth/pymdea.git /example/file/path/pymdea.git

Create environment¶

Navigate to the cloned repository directory:

cd /example/file/path/pymdea.git

and run the uv command:

uv sync

This will create a python virtual environment in a subdirectory called .../pymdea.git/.venv which contains a python executable and all packages necessary to run pymdea. The command will display text output in the command line, which should look like this:

❯ uv sync
Using Python 3.12.5
Creating virtualenv at: .venv
Resolved 107 packages in 1.01s
   Built pymdea @ file:///D:/repos/pymdea.git
Prepared 101 packages in 819ms
░░░░░░░░░░░░░░░░░░░░ [0/101] Installing
Installed 101 packages in 5.05s
 + asttokens==2.4.1
 + attrs==24.2.0
 + babel==2.16.0
 ... (all 101 packages uv installs will be listed out)

Note: the file path, D:/repos/pymdea.git, is the directory on my own file system my clone of pymdea is located.

Verify integrity¶

To confirm the files work properly, run the uv command:

uv run pytest

This runs the pymdea test suite from test_core.py and display a summary of the results in the command line output.

With these steps complete, you're now ready to import the pymdea.core and pymdea.plot modules and work with them!

Loading pymdea¶

Import the pymdea modules as you would any other Python package.

In [1]:

from pymdea.core import DeaEngine, DeaLoader
from pymdea.plot import DeaPlotter

from pymdea.core import DeaEngine, DeaLoader from pymdea.plot import DeaPlotter

Loading data¶

The pymdea.core module provides a DeaLoader class with helper methods for wrangling data. It also has a make_sample_data method for generating sample data for testing purposes, which is what this notebook will use to illustrate an example workflow.

In [2]:

dea_loader = DeaLoader()
dea_loader.make_sample_data(30000)

dea_loader = DeaLoader() dea_loader.make_sample_data(30000)

Out[2]:

<pymdea.core.DeaLoader at 0x296be3bf8f0>

Analyzing data¶

The DeaEngine class from pymdea.core contains the methods for analyzing data once it's been loaded. Pass DeaLoader to DeaEngine to pass the data from the loader to the engine, then invoke the analyze method to run an analysis.

In [3]:

dea_engine = DeaEngine(dea_loader)
dea_engine.analyze_with_stripes(fit_start=0.1, fit_stop=0.9, n_stripes=60)

dea_engine = DeaEngine(dea_loader) dea_engine.analyze_with_stripes(fit_start=0.1, fit_stop=0.9, n_stripes=60)

100%|██████████| 582/582 [00:00<00:00, 1232.44it/s]

--------------------------------- 
result 
 δ: 0.5078088643108836 
 μ (rule 1): 1.5078088643108836 
 μ (rule 2): 2.9692448680608186 
---------------------------------

Viewing results¶

The DeaPlotter class from pymdea.plot contains methods for viewing the results and producing plots. Pass DeaEngine to DeaPlotter to pass the results from the engine to the plotter, then invoke the plotting methods to view different summaries of the results.

In [4]:

dea_plot = DeaPlotter(dea_engine)

dea_plot = DeaPlotter(dea_engine)

The most illustrative figure is plotting the Shannon entropy against the log of the window lengths, S(L) vs ln(L):

In [5]:

dea_plot.s_vs_l()

dea_plot.s_vs_l()

It's also useful to quickly compare the results of the different rules for computing the mu parameter:

In [6]:

dea_plot.mu_candidates()

dea_plot.mu_candidates()