Make compatible with pip packaging system

README.md

@@ -20,11 +20,15 @@ Both methods are illustrated in the following figures :
 ![Illustration of sample based scoring.](https://user-images.githubusercontent.com/747240/244666630-cdfe12cc-22a2-4b23-be15-3e60dbedb437.png)
 ![Illustration of event based scoring.](https://user-images.githubusercontent.com/747240/244666619-8dd90008-79af-4836-8769-daa204bbe16c.png)
 
-## Code
+## Installation
+
+The package can be installed through pip using the following command :
 
-An example usage of the library is provided in `example.ipynb`.
+`pip install timescoring`
+
+## Code
 
-The library provides three classes :
+The `timescoring` package provides three classes :
 
 - `annotation.Annotation` : store annotations
 - `scoring.SampleScoring(ref, hyp)` : Compute sample based scoring
@@ -58,7 +62,7 @@ Scores are provided as attributes of the scoring class. The following metrics ca
 ```python
 ## Loading Annotations ##
 
-from annotations import Annotation
+from timescoring.annotations import Annotation
 
 # Annotation objects can be instantiated from a binary mask
 
@@ -82,15 +86,15 @@ labels = Annotation(events, fs, numSamples)
 
 ## Computing performance score ## 
 
-import scoring
-import visualization
+from timescoring import scoring
+from timescoring import visualization
 
 fs = 1
 duration = 66*60
 ref = Annotation([(8*60, 12*60), (30*60, 35*60), (48*60, 50*60)], fs, duration)
 hyp = Annotation([(8*60, 12*60), (28*60, 32*60), (50.5*60, 51*60), (60*60, 62*60)], fs, duration)
 scores = scoring.SampleScoring(ref, hyp)
-figSamples = visualization.plotSampleScoring(ref, hyp, param)
+figSamples = visualization.plotSampleScoring(ref, hyp)
 
 # Scores can also be computed per event
 param = scoring.EventScoring.Parameters(

pyproject.toml

@@ -0,0 +1,28 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "timescoring"
+version = "0.0.1"
+authors = [
+  { name="Jonathan Dan", email="[email protected]" },
+  { name="Una Pale", email="[email protected]" },
+  { name="PEDESITE" }
+]
+description = "Library for measuring performance of time series classification"
+readme = "README.md"
+requires-python = ">=3.10"
+dependencies = [
+  "matplotlib>=3.7.1",
+  "numpy>=1.24.3 ",
+  "nptyping>=2.5.0"]
+classifiers = [
+    "Programming Language :: Python :: 3",
+    "License :: OSI Approved :: MIT License",
+    "Operating System :: OS Independent",
+]
+
+[project.urls]
+"Homepage" = "https://github.com/esl-epfl/epilepsy_performance_metrics"
+"Bug Tracker" = "https://github.com/esl-epfl/epilepsy_performance_metrics/issues"

scoring.py → src/timescoring/scoring.py

@@ -6,7 +6,7 @@
 
 import numpy as np
 
-from annotations import Annotation
+from .annotations import Annotation
 
 class _Scoring:
     """" Base class for different scoring methods. The class provides the common

visualization.py → src/timescoring/visualization.py

@@ -1,8 +1,8 @@
 import matplotlib.pyplot as plt
 import numpy as np
 
-from annotations import Annotation
-import scoring
+from .annotations import Annotation
+from . import scoring
 
 def plotSampleScoring(ref : Annotation, hyp : Annotation, fs : int = 1) -> plt.figure:
     """Build an overview plot showing the outcome of sample scoring.

test.py → tests/test.py

@@ -8,8 +8,8 @@
 
 import numpy as np
 
-from annotations import Annotation
-import scoring 
+from src.timescoring.annotations import Annotation
+from src.timescoring import scoring 
 
 class TestAnnotation(unittest.TestCase):
     def assertListOfTupleEqual(expected, actual, message):