Documentation and formatting

README.md

@@ -13,7 +13,8 @@ and the Flutter guide for
 
 [![Dart CI](https://github.com/XPerianer/Bandart/actions/workflows/test-package.yml/badge.svg)](https://github.com/XPerianer/Bandart/actions/workflows/test-package.yml)
 
-Bandart (combination of "Bandits" and "Dart") is a library for bandit algorithms in Dart. It provides Bayesian Models for statistical analysis for the data.
+Bandart (combination of "Bandits" and "Dart") is a library for bandit algorithms in Dart.
+It provides Bayesian Models for statistical analysis for the data.
 
 ## Features
 - Bandit algorithms:
@@ -24,6 +25,7 @@ Bandart (combination of "Bandits" and "Dart") is a library for bandit algorithms
   - Gaussian Model: Approximating mean and variance using conjugate priors
 
 
+
 ## Getting started
 
 Install the package using ```dart pub add bandart```.
@@ -63,7 +65,6 @@ gaussianModel.history = history
 // Update the samples
 gaussianModel.sample()
 
-// Look at the results
 print(gaussianModel.maxProbabilities())
 // Prints around [0.2, 0.8]
 ```
@@ -82,8 +83,8 @@ var gaussianModel = GaussianModel(
 
 var policy = ThompsonSampling(numberOfInterventions: 2);
 
-// Prints either 0 or 1 (randomized), but will more often pick 1 as this is the intervention with the better history
 print(policy.choseAction({}, history));
+// Prints either 0 or 1 (randomized), but will more often pick 1 as this is the intervention with the better history
 
 ```
 
@@ -100,9 +101,4 @@ To update the mock code, run ```dart run build_runner build```.
 - Performance: Since Dart is most often used for mobile development, the goal of the library is to support calculations fast enough to run on smartphones.
 - Seedability: When seeded with the same seed, the library should always return the same result. 
 - Well tested
-- Extensible
-- 
 
-TODO: Tell users more about the package: where to find more information, how to
-contribute to the package, how to file issues, what response they can expect
-from the package authors, and more.

lib/bandart.dart

@@ -1,3 +1,4 @@
+/// Library for Bandit Algorithms with underlying Bayesian Models
 library bandart;
 
 export 'src/dataframe.dart' show DataFrame;
@@ -9,4 +10,4 @@ export 'src/models/gaussian.dart' show GaussianModel;
 
 export 'src/policies/policy.dart' show Policy;
 export 'src/policies/fixed_policy.dart' show FixedPolicy;
-export 'src/policies/thompson_sampling.dart' show ThompsonSampling;
+export 'src/policies/thompson_sampling.dart' show ThompsonSampling;

lib/helpers.dart

@@ -1,3 +1,4 @@
+/// Collections of helpful functions to write easier code.
 library helpers;
 
 export 'src/helpers.dart';

lib/src/helpers.dart

@@ -4,6 +4,7 @@ import 'dart:math';
 
 import 'package:collection/collection.dart';
 
+/// Returns the index of the largest element in the list
 int argMax<T extends num>(List<T> list) {
   return list
       .asMap()

lib/src/policies/thompson_sampling.dart

@@ -3,6 +3,9 @@ import 'dart:math';
 import 'package:bandart/bandart.dart';
 import 'package:bandart/helpers.dart';
 
+/// ThompsonSampling is a policy that picks an intervention based on the probability of it being the best intervention.
+///
+/// For example: If the probability of intervention 0 beeing the best intervention is 0.8, then ThompsonSampling will pick intervention 0 with a probability of 0.8.
 class ThompsonSampling implements Policy {
   final SamplingModel model;
   final Random random;