{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "31a86ad2-5ed8-42b8-ad15-7831af7e0316",
   "metadata": {},
   "source": [
    "# Learners\n",
    "Coba Learners choose which actions to take in a context and learn from results of those actions. The Learner interface is:\n",
    "\n",
    "```python\n",
    "class Learner:\n",
    "    def score(self, context: Context, actions: Sequence[Action], action:Action) -> float:\n",
    "        \"\"\"(Optional) Return the propensity score for the given action.\"\"\"\n",
    "\n",
    "    def predict(self, context: Context, actions: Sequence[Action]) -> Action | Tuple[Action,float]:\n",
    "        \"\"\"Choose which action to take.\"\"\"\n",
    "\n",
    "    def learn(self, context: Context, action: Action, reward: Reward, probability: float, **kwargs) -> None:\n",
    "        \"\"\"Learn from the results of an action in a context.\"\"\"\"\n",
    "```\n",
    "\n",
    "The `Context` and `Action` type hints are described in more detail in the [Interactions](Interactions.ipynb) notebook."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b9606dd7-c351-4404-bd75-a553b6c5ac6f",
   "metadata": {},
   "source": [
    "## Minimal Viable Learner\n",
    "\n",
    "To create a custom on-policy learner we implement two methods in the learner interface: `predict` and `learn`. \n",
    "\n",
    "As an example here we create a learner that randomly chooses actions without learning anything.\n",
    "\n",
    "In this example we use Coba's built-in randomization but the random choosing could be implemented using any method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "50b90530-dce3-47d2-987b-0eb7510c6770",
   "metadata": {},
   "outputs": [],
   "source": [
    "import coba as cb\n",
    "\n",
    "class MyRandomLearner:\n",
    "    def predict(self, context, actions):\n",
    "        return cb.random.choice(actions)\n",
    "    def learn(self, context, action, reward, probability):\n",
    "        pass"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c22add42-c434-444f-8eef-c39bccb86af7",
   "metadata": {},
   "source": [
    "And that is all, we can now use this learner in an experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "559ad28d-1e9d-4f3d-985a-22af224fba04",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2024-02-08 14:30:23 -- Experiment Started\n",
      "2024-02-08 14:30:23 -- Recording Learner 0 parameters... (0.0 seconds) (completed)\n",
      "2024-02-08 14:30:23 -- Recording Evaluator 0 parameters... (0.0 seconds) (completed)\n",
      "2024-02-08 14:30:23 -- Peeking at Environment 0... (0.0 seconds) (completed)\n",
      "2024-02-08 14:30:23 -- Recording Environment 0 parameters... (0.0 seconds) (completed)\n",
      "2024-02-08 14:30:23 -- Evaluating Learner 0 on Environment 0... (0.02 seconds) (completed)\n",
      "2024-02-08 14:30:23 -- Experiment Finished\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cb.random.seed(1)\n",
    "env = cb.Environments.from_linear_synthetic(100)\n",
    "lrn = MyRandomLearner()\n",
    "cb.Experiment(env,lrn).run().plot_learners()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "99cb44ce-cbb0-4442-81cd-4e00dcfe13e1",
   "metadata": {},
   "source": [
    "## Epsilon Bandit Learner\n",
    "\n",
    "Of course we don't want to just randomly pick actions so here we implement an a bandit learner using epsilon-greedy exploration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "830fe483-17ec-44ca-b841-b5d50e2d1654",
   "metadata": {},
   "outputs": [],
   "source": [
    "import coba as cb\n",
    "\n",
    "class MyEpsilonBanditLearner:\n",
    "    def __init__(self, epsilon,seed=1):\n",
    "        self._epsilon = epsilon\n",
    "        self._means   = {}\n",
    "        self._counts  = {}\n",
    "        self._rng     = cb.CobaRandom(seed)\n",
    "    \n",
    "    def predict(self, context, actions):\n",
    "        if set(actions) - self._means.keys():\n",
    "            #unseen\n",
    "            return next(iter(actions-self._means.keys()))\n",
    "        elif self._rng.random() <= self._epsilon: \n",
    "            #explore\n",
    "            return self._rng.choice(actions)\n",
    "        else:\n",
    "            #greed\n",
    "            return max(actions,key=self._means.__getitem__)\n",
    "    \n",
    "    def learn(self, context, action, reward, probability):\n",
    "        n,m = self._counts.get(action,0),self._means.get(action,0)\n",
    "        self._counts[action] = n + 1\n",
    "        self._means[action]  = m + (reward-m)/(n+1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a8aebdec-10de-4230-8b54-7c8b39e18f0a",
   "metadata": {},
   "source": [
    "We can now compare the performance of our epsilon bandit learner to our random learner"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "37765fce-8d60-49af-9def-fb91d0c7caca",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2024-02-08 14:31:22 -- Experiment Started\n",
      "2024-02-08 14:31:22 -- Recording Learner 0 parameters... (0.0 seconds) (completed)\n",
      "2024-02-08 14:31:22 -- Recording Evaluator 0 parameters... (0.0 seconds) (completed)\n",
      "2024-02-08 14:31:22 -- Recording Learner 1 parameters... (0.0 seconds) (completed)\n",
      "2024-02-08 14:31:22 -- Peeking at Environment 0... (0.0 seconds) (completed)\n",
      "2024-02-08 14:31:22 -- Recording Environment 0 parameters... (0.0 seconds) (completed)\n",
      "2024-02-08 14:31:23 -- Evaluating Learner 0 on Environment 0... (0.0 seconds) (completed)\n",
      "2024-02-08 14:31:23 -- Evaluating Learner 1 on Environment 0... (0.0 seconds) (completed)\n",
      "2024-02-08 14:31:23 -- Experiment Finished\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cb.random.seed(4)\n",
    "env = cb.Environments.from_linear_synthetic(100,n_context_features=0,n_action_features=0)\n",
    "lrn = [MyEpsilonBanditLearner(.1), MyRandomLearner()]\n",
    "cb.Experiment(env,lrn).run().plot_learners()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fc73db84-3398-4fe6-9bae-90202b88a05f",
   "metadata": {},
   "source": [
    "## Learner Parameters\n",
    "\n",
    "To understand what makes our learner work it is useful to keep track of learner parameters. Here we add to our bandit learner."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "0dc9ec05-b41b-4dd5-969b-49555ee6d5df",
   "metadata": {},
   "outputs": [],
   "source": [
    "import coba as cb\n",
    "#The super class MyEpsilonBanditLearner is defined above\n",
    "class MyEpsilonBanditLearnerWithParams(MyEpsilonBanditLearner):\n",
    "    @property\n",
    "    def params(self):\n",
    "        return {'epsilon': self._epsilon}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "0750ebf8-e8b3-43b6-b7a4-e91048ce5a12",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "env = cb.Environments.from_linear_synthetic(10_000,n_context_features=0,n_action_features=0,seed=range(10)).noise(reward=(0,.5)).binary()\n",
    "lrn = [ MyEpsilonBanditLearnerWithParams(e) for e in [0,.01,.1,.5] ]\n",
    "cb.Experiment(env,lrn).run(quiet=True,processes=8).plot_learners()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6a3e4322-0099-4c7a-ab87-db1c0b5ef3a6",
   "metadata": {},
   "source": [
    "## Recording Probabilities\n",
    "\n",
    "It can be useful to keep track of the probability of taking actions for this use case probabilities can be returned with actions.\n",
    "\n",
    "When probabilities are returned from `predict` the probability will be passed to `learn` as the `probability` arg."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "cc796349-d66f-4012-b7d4-269634627ba9",
   "metadata": {},
   "outputs": [],
   "source": [
    "import coba as cb\n",
    "\n",
    "class MyRandomLearnerWithProbabilities:\n",
    "    def predict(self, context, actions):\n",
    "        probability = 1/len(actions)\n",
    "        chosen_act  = cb.random.choice(actions)\n",
    "        return chosen_act, probability\n",
    "        \n",
    "    def learn(self, context, action, reward, probability):\n",
    "        #When this is called probability equals the probability returned from predict\n",
    "        pass"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a585302e-e115-4581-9e9c-7c4db5470366",
   "metadata": {},
   "source": [
    "See below that we have the probabilities from our learner stored in result."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "5ed0f6c0-584f-4219-b6c1-8eb2f9c8d0f4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>environment_id</th>\n",
       "      <th>learner_id</th>\n",
       "      <th>evaluator_id</th>\n",
       "      <th>index</th>\n",
       "      <th>action</th>\n",
       "      <th>probability</th>\n",
       "      <th>reward</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>[0, 1, 0, 0, 0]</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.52699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>[0, 1, 0, 0, 0]</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.56189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>[0, 1, 0, 0, 0]</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.48935</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>[1, 0, 0, 0, 0]</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.53100</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   environment_id  learner_id  evaluator_id  index           action  \\\n",
       "0               0           0             0      1  [0, 1, 0, 0, 0]   \n",
       "1               0           0             0      2  [0, 1, 0, 0, 0]   \n",
       "2               0           0             0      3  [0, 1, 0, 0, 0]   \n",
       "3               0           0             0      4  [1, 0, 0, 0, 0]   \n",
       "\n",
       "   probability   reward  \n",
       "0          0.2  0.52699  \n",
       "1          0.2  0.56189  \n",
       "2          0.2  0.48935  \n",
       "3          0.2  0.53100  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "env = cb.Environments.from_linear_synthetic(4,n_action_features=0)\n",
    "lrn = MyRandomLearnerWithProbabilities()\n",
    "result = cb.Experiment(env,lrn).run(quiet=True)\n",
    "result.interactions.to_pandas()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84c3f7d7-873d-4245-825c-5dc5f313b51b",
   "metadata": {},
   "source": [
    "## Custom Arguments\n",
    "\n",
    "Some learners require additional information to be shared between `predict` and `learn`. \n",
    "\n",
    "For this use case we support explicitly returning kwargs.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "7ac1be13-f128-49bb-aa1e-4000ff186ca0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import coba as cb\n",
    "\n",
    "class MyRandomLearnerWithKwargs:\n",
    "\n",
    "    def predict(self, context, actions):\n",
    "        probability = 1/len(actions)\n",
    "        chosen_act  = cb.random.choice(actions)\n",
    "        kwargs      = {'kwarg1' : 1, 'item1': 2, 'b': 3}\n",
    "        return chosen_act, probability, kwargs\n",
    "\n",
    "    def learn(self, context, action, reward, probability, kwarg1, item1, b):\n",
    "        pass"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d78b1587-0eb0-4e78-a146-54137d2c936d",
   "metadata": {},
   "source": [
    "## Score Method\n",
    "\n",
    "The score method is largely optional. When it is implemented it used in the following ways:\n",
    "1. It is optional to reduce the variance of IPS off-policy evaluation for discrete action spaces.\n",
    "2. It is required to perform IPS off-policy evaluation in continuous action spaces.\n",
    "3. It is required to use the RejectionCB evaluator due to rejection sampling."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "3e173e29-f8ac-4c94-a6af-ed5c5a4b5ccb",
   "metadata": {},
   "outputs": [],
   "source": [
    "import coba as cb\n",
    "\n",
    "class MyRandomLearnerWithProbabilities:\n",
    "\n",
    "    def score(self, context, actions, action):\n",
    "        #this should return the probability of taking `action`\n",
    "        #given context and actions. In discrete settings action\n",
    "        #should always be in actions.\n",
    "        return 1/len(actions)\n",
    "        \n",
    "    def predict(self, context, actions):\n",
    "        probability = 1/len(actions)\n",
    "        chosen_act  = cb.random.choice(actions)\n",
    "        return chosen_act, probability\n",
    "\n",
    "    def learn(self, context, action, reward, probability):\n",
    "        pass"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d0ecb937-bb3a-487c-9fd5-f794a1c50b67",
   "metadata": {},
   "source": [
    "## Advanced Metrics\n",
    "\n",
    "Coba's built-in evaluators can track all desired metric for learners using the dictionary at `CobaContext.learning_info`.\n",
    "\n",
    "Below we show how to track the current estimated mean for each action as a bandit learner gets observations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "5114a10c-b9f6-4bb3-a3d0-048718dca48f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import coba as cb\n",
    "\n",
    "#The super class MyEpsilonBanditLearner is defined above\n",
    "class MyEpsilonBanditLearnerWithMetrics(MyEpsilonBanditLearner):\n",
    "    def learn(self,*args):\n",
    "        super().learn(*args)\n",
    "        cb.CobaContext.learning_info.update(self._means)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "d6494e6c-5e81-422a-8414-3d00f3595eb9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>environment_id</th>\n",
       "      <th>learner_id</th>\n",
       "      <th>evaluator_id</th>\n",
       "      <th>index</th>\n",
       "      <th>(0, 1)</th>\n",
       "      <th>(1, 0)</th>\n",
       "      <th>action</th>\n",
       "      <th>reward</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>[1, 0]</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>[0, 1]</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>[0, 1]</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>0.50000</td>\n",
       "      <td>[1, 0]</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0.66667</td>\n",
       "      <td>0.50000</td>\n",
       "      <td>[0, 1]</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.66667</td>\n",
       "      <td>0.33333</td>\n",
       "      <td>[1, 0]</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>0.75000</td>\n",
       "      <td>0.33333</td>\n",
       "      <td>[0, 1]</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>0.75000</td>\n",
       "      <td>0.50000</td>\n",
       "      <td>[1, 0]</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   environment_id  learner_id  evaluator_id  index   (0, 1)   (1, 0)  action  \\\n",
       "0               0           0             0      1      NaN  1.00000  [1, 0]   \n",
       "1               0           0             0      2  1.00000  1.00000  [0, 1]   \n",
       "2               0           0             0      3  1.00000  1.00000  [0, 1]   \n",
       "3               0           0             0      4  1.00000  0.50000  [1, 0]   \n",
       "4               0           0             0      5  0.66667  0.50000  [0, 1]   \n",
       "5               0           0             0      6  0.66667  0.33333  [1, 0]   \n",
       "6               0           0             0      7  0.75000  0.33333  [0, 1]   \n",
       "7               0           0             0      8  0.75000  0.50000  [1, 0]   \n",
       "\n",
       "   reward  \n",
       "0       1  \n",
       "1       1  \n",
       "2       1  \n",
       "3       0  \n",
       "4       0  \n",
       "5       0  \n",
       "6       1  \n",
       "7       1  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "env = cb.Environments.from_linear_synthetic(8,n_context_features=0,n_action_features=0,n_actions=2).noise(reward=(0,0.5)).binary()\n",
    "lrn = MyEpsilonBanditLearnerWithMetrics(.5)\n",
    "result = cb.Experiment(env,lrn).run(quiet=True)\n",
    "result.interactions.to_pandas()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}