Metadata-Version: 2.1
Name: optunacy
Version: 0.3.0
Summary: Optunacy is a contour plot generator for Optuna studies.
Home-page: https://github.com/retospect/optunacy
License: GPL-3
Keywords: optuna,contour,plot,snn,hyperparameter,optimization
Author: Reto Stamm
Maintainer: Reto Stamm
Requires-Python: >=3.9,<4.0
Classifier: License :: Other/Proprietary License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Requires-Dist: matplotlib (>=3.8.2,<4.0.0)
Requires-Dist: numpy (>=1.26.2,<2.0.0)
Requires-Dist: optuna (>=3.4.0,<4.0.0)
Requires-Dist: plotly (>=5.18.0,<6.0.0)
Requires-Dist: scipy (>=1.11.4,<2.0.0)
Project-URL: Repository, https://github.com/retospect/optunacy
Description-Content-Type: text/markdown

[![check](https://github.com/retospect/optunacy/actions/workflows/check.yml/badge.svg)](https://github.com/retospect/optunacy/actions/workflows/check.yml)
# Optunacy - a better Optuna contour plotter

Plots Optuna contour plots trivially and with detailed mouseovers.
It collects data from all the hashes in Optuna, as well as the Objectives if you provide them.

The plot function has these options:

``` python
def plot(self, 
         x_name, # string
         y_name, # string
         z_name=None,  # string
         x_range=None, # range, ex: x_range=(3,9)
         y_range=None, # range
         z_clip=None,  # Clamps colors to max and min
         interpol="linear"): # <a href="https://docs.scipy.org/doc/scipy/reference/generated/scipy.interpolate.griddata.html">interpolation method</a>
```

## Example

```python
def optuna_objective(trial, train_loader, test_loader):
    ...
    trial.set_user_attr("Neurons per Hidden Layer", nphl)
    ...
    # The objectives have a printable name and direction
    return             validation_accuracy,   spikes_per_digit
objective_names      = ["Validation Accuracy", "Spikes per Digit"]
objective_directions = ["maximize",            "minimize"]

study = optuna.create_study(study_name="Minimize spikes, maximize accuracy",
                            directions=objective_directions)

study.optimize( lambda trial:
                optuna_objective(trial, train_loader, test_loader),
                n_trials=100)

see = OPlot(study, objective_names)
see.plot("Hidden Layers", "Neurons per Hidden Layer", "Validation Accuracy", z_clip=(0.8,1))
```
![Sample output](imgs/sample1.png)

## Installation

```pip install optunacy```

