A simple entrypoint to perform a trend analysis in Python.
Repository
PyPI
Documentation
pyEasyTrend can be installed using pip:
pip install pyEasyTrendor downloading / cloning the repository and, from the root folder of the project, running:
python setup.py installTo update the package via pip, you can use:
pip install --user --upgrade pyeasytrendYou can check the full documentation online. Alternatively, a copy of the documentation is available in the docs folder
import pyeasytrend #import the library
#Load some sample data and create a dataframe
data = {'StudyTime':[24, 44, 21, 45, 54, 26, 57, 34, 33, 12, 17, 21, 58, 41, 29, 55, 42, 40, 21, 9, 39, 30, 17, 31, 51, 42, 30, 3, 20, 21, 4, 16, 26, 6, 18, 50, 60, 13, 23, 13, 3, 35, 38, 51, 12, 35, 7, 42, 20, 41, 37, 56, 19, 57, 12, 49, 15, 6, 43, 7, 40, 12, 35, 4, 46, 29, 6, 38, 36, 33, 21, 33, 50, 54, 25, 38, 48, 17, 28, 48, 16, 50, 24, 15, 40, 54, 40, 42, 2, 20, 24, 21, 37, 15, 52, 36, 5, 7, 29, 21],
'Score':[18, 47, 21, 60, 80, 18, 100, 28, 41, 7, 12, 17, 82, 45, 33, 94, 41, 55, 9, 6, 53, 24, 13, 35, 62, 43, 33, 2, 17, 10, 0, 7, 14, 0, 14, 72, 94, 7, 14, 3, 0, 43, 39, 80, 5, 39, 4, 43, 14, 37, 39, 80, 16, 94, 7, 55, 13, 2, 45, 6, 55, 7, 35, 0, 69, 18, 0, 45, 43, 27, 11, 37, 67, 82, 16, 41, 74, 10, 19, 55, 14, 60, 18, 7, 55, 64, 37, 60, 2, 10, 17, 14, 30, 6, 69, 32, 2, 1, 32, 10]}
df = pd.DataFrame(data)
#Analyze the data using up to a Quartic model (y = ax^4 bx^3 cx^2 dx q, maxDegree = 4), and generate a visual representation of the analysis (visualize = True)
results = pyeasytrend.analyzeTrend(df.StudyTime, df.Score, maxDegree=4, visualize=True)
#Put the results in a pandas Table
pyeasytrend.tablifyResults(results)
| Order | R2 | SSE | F | pvalue | AIC | BIC |
|---|---|---|---|---|---|---|
| 1 | 0.907946 | 6498.52 | 966.59 | 1.11022e-16 | 703.204 | 705.809 |
| 2 | 0.952677 | 3340.75 | 91.6874 | 1.11022e-15 | 638.666 | 643.876 |
| 3 | 0.952691 | 3339.76 | 0.0283989 | 0.866529 | 640.636 | 648.452 |
This and other examples can also be found in the Tutorial folder in the form of Jupyter Notebook.
- Numpy
- Scipy
- Pandas
- Matplotlib
Feel free to contact me for questions, suggestions or to give me advice as well at: [email protected]
