From d5b00e046302f0ff7e77783573ac340088404bb7 Mon Sep 17 00:00:00 2001 From: usaito Date: Sun, 8 Nov 2020 01:43:32 +0900 Subject: [PATCH 1/4] add obp.dataset.multiclass --- docs/obp.rst | 1 + 1 file changed, 1 insertion(+) diff --git a/docs/obp.rst b/docs/obp.rst index 05fee0ad..0ed028a5 100644 --- a/docs/obp.rst +++ b/docs/obp.rst @@ -31,6 +31,7 @@ dataset module obp.dataset.base obp.dataset.real obp.dataset.synthetic + obp.dataset.multiclass simulator module From 09d848477c797e432e24540a0fe1af391185fc86 Mon Sep 17 00:00:00 2001 From: usaito Date: Sun, 8 Nov 2020 01:43:52 +0900 Subject: [PATCH 2/4] rerun with a new version --- examples/quickstart/quickstart.ipynb | 65 ++++++++++--------- .../quickstart/quickstart_synthetic.ipynb | 14 ++-- 2 files changed, 41 insertions(+), 38 deletions(-) diff --git a/examples/quickstart/quickstart.ipynb b/examples/quickstart/quickstart.ipynb index 497498ee..288d8597 100644 --- a/examples/quickstart/quickstart.ipynb +++ b/examples/quickstart/quickstart.ipynb @@ -44,7 +44,7 @@ "output_type": "stream", "name": "stdout", "text": [ - "0.3.1\n" + "0.3.2\n" ] } ], @@ -73,17 +73,20 @@ }, { "cell_type": "code", - "execution_count": 3, + "execution_count": 4, "metadata": { "tags": [] }, "outputs": [ { - "output_type": "stream", - "name": "stdout", - "text": [ - "dict_keys(['n_rounds', 'n_actions', 'action', 'position', 'reward', 'reward_test', 'pscore', 'context', 'action_context'])\n" - ] + "output_type": "execute_result", + "data": { + "text/plain": [ + "dict_keys(['n_rounds', 'n_actions', 'action', 'position', 'reward', 'reward_test', 'pscore', 'context', 'action_context'])" + ] + }, + "metadata": {}, + "execution_count": 4 } ], "source": [ @@ -96,7 +99,7 @@ "bandit_feedback = dataset.obtain_batch_bandit_feedback()\n", "\n", "# `bandit_feedback` is a dictionary storing logged bandit feedback\n", - "print(bandit_feedback.keys())" + "bandit_feedback.keys()" ] }, { @@ -108,7 +111,7 @@ }, { "cell_type": "code", - "execution_count": 4, + "execution_count": 5, "metadata": {}, "outputs": [ { @@ -119,7 +122,7 @@ ] }, "metadata": {}, - "execution_count": 4 + "execution_count": 5 } ], "source": [ @@ -129,7 +132,7 @@ }, { "cell_type": "code", - "execution_count": 5, + "execution_count": 6, "metadata": {}, "outputs": [ { @@ -140,7 +143,7 @@ ] }, "metadata": {}, - "execution_count": 5 + "execution_count": 6 } ], "source": [ @@ -150,7 +153,7 @@ }, { "cell_type": "code", - "execution_count": 6, + "execution_count": 7, "metadata": {}, "outputs": [ { @@ -161,7 +164,7 @@ ] }, "metadata": {}, - "execution_count": 6 + "execution_count": 7 } ], "source": [ @@ -171,7 +174,7 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 8, "metadata": {}, "outputs": [ { @@ -182,7 +185,7 @@ ] }, "metadata": {}, - "execution_count": 7 + "execution_count": 8 } ], "source": [ @@ -192,7 +195,7 @@ }, { "cell_type": "code", - "execution_count": 8, + "execution_count": 9, "metadata": {}, "outputs": [ { @@ -203,7 +206,7 @@ ] }, "metadata": {}, - "execution_count": 8 + "execution_count": 9 } ], "source": [ @@ -234,7 +237,7 @@ }, { "cell_type": "code", - "execution_count": 9, + "execution_count": 10, "metadata": { "tags": [] }, @@ -295,7 +298,7 @@ ] }, "metadata": {}, - "execution_count": 9 + "execution_count": 10 } ], "source": [ @@ -340,7 +343,7 @@ }, { "cell_type": "code", - "execution_count": 10, + "execution_count": 11, "metadata": {}, "outputs": [], "source": [ @@ -365,7 +368,7 @@ }, { "cell_type": "code", - "execution_count": 11, + "execution_count": 12, "metadata": { "tags": [] }, @@ -388,7 +391,7 @@ }, { "cell_type": "code", - "execution_count": 12, + "execution_count": 13, "metadata": {}, "outputs": [], "source": [ @@ -404,7 +407,7 @@ }, { "cell_type": "code", - "execution_count": 13, + "execution_count": 14, "metadata": {}, "outputs": [ { @@ -419,7 +422,7 @@ "text/html": "
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
estimated_policy_valuerelative_estimated_policy_value
ipw0.0045531.198126
dm0.0034040.895800
dr0.0046511.224077
\n
" }, "metadata": {}, - "execution_count": 13 + "execution_count": 14 } ], "source": [ @@ -429,7 +432,7 @@ }, { "cell_type": "code", - "execution_count": 14, + "execution_count": 15, "metadata": {}, "outputs": [ { @@ -444,7 +447,7 @@ "text/html": "
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
mean95.0% CI (lower)95.0% CI (upper)
ipw0.0045400.0015690.009069
dm0.0034070.0033660.003445
dr0.0046400.0016250.009240
\n
" }, "metadata": {}, - "execution_count": 14 + "execution_count": 15 } ], "source": [ @@ -456,14 +459,14 @@ }, { "cell_type": "code", - "execution_count": 15, + "execution_count": 16, "metadata": {}, "outputs": [ { "output_type": "display_data", "data": { "text/plain": "
", - "image/svg+xml": "\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n", + "image/svg+xml": "\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n", "image/png": "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\n" }, "metadata": {} @@ -481,14 +484,14 @@ }, { "cell_type": "code", - "execution_count": 16, + "execution_count": 17, "metadata": {}, "outputs": [ { "output_type": "display_data", "data": { "text/plain": "
", - "image/svg+xml": "\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n", + "image/svg+xml": "\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n", "image/png": "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\n" }, "metadata": {} diff --git a/examples/quickstart/quickstart_synthetic.ipynb b/examples/quickstart/quickstart_synthetic.ipynb index 36cad41f..ae6469d4 100644 --- a/examples/quickstart/quickstart_synthetic.ipynb +++ b/examples/quickstart/quickstart_synthetic.ipynb @@ -75,7 +75,7 @@ "output_type": "stream", "name": "stdout", "text": [ - "0.3.1\n" + "0.3.2\n" ] } ], @@ -333,14 +333,14 @@ "output_type": "stream", "name": "stdout", "text": [ - " mean 95.0% CI (lower) 95.0% CI (upper)\nipw 0.632349 0.628700 0.636076\ndm 0.612067 0.610717 0.613459\ndr 0.632265 0.628844 0.636009 \n\n" + " mean 95.0% CI (lower) 95.0% CI (upper)\nipw 0.632001 0.628205 0.635098\ndm 0.612102 0.611062 0.613282\ndr 0.632011 0.628249 0.635968 \n\n" ] }, { "output_type": "display_data", "data": { "text/plain": "
", - "image/svg+xml": "\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n", + "image/svg+xml": "\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n", "image/png": "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\n" }, "metadata": {} @@ -373,14 +373,14 @@ "output_type": "stream", "name": "stdout", "text": [ - " mean 95.0% CI (lower) 95.0% CI (upper)\nipw 0.635290 0.631430 0.639023\ndm 0.611352 0.609821 0.612917\ndr 0.634975 0.631641 0.639456 \n\n" + " mean 95.0% CI (lower) 95.0% CI (upper)\nipw 0.635143 0.631193 0.639453\ndm 0.611249 0.609818 0.612619\ndr 0.635139 0.631754 0.638721 \n\n" ] }, { "output_type": "display_data", "data": { "text/plain": "
", - "image/svg+xml": "\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n", + "image/svg+xml": "\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n", "image/png": "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\n" }, "metadata": {} @@ -413,14 +413,14 @@ "output_type": "stream", "name": "stdout", "text": [ - " mean 95.0% CI (lower) 95.0% CI (upper)\nipw 0.606584 0.603318 0.610094\ndm 0.606224 0.604829 0.607496\ndr 0.606971 0.603938 0.609665 \n\n" + " mean 95.0% CI (lower) 95.0% CI (upper)\nipw 0.606567 0.603520 0.609491\ndm 0.606244 0.604977 0.607734\ndr 0.606968 0.603830 0.609725 \n\n" ] }, { "output_type": "display_data", "data": { "text/plain": "
", - "image/svg+xml": "\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n", + "image/svg+xml": "\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n", "image/png": "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\n" }, "metadata": {} From b5a3a66c8bb8321cb853e653d27aea3fa265ebc6 Mon Sep 17 00:00:00 2001 From: usaito Date: Sun, 8 Nov 2020 01:44:48 +0900 Subject: [PATCH 3/4] fix some docstring inconsistencies --- obp/dataset/multiclass.py | 24 ++++++++++--------- obp/dataset/real.py | 14 +++++------ obp/dataset/synthetic.py | 18 +++++++-------- obp/ope/estimators.py | 46 ++++++++++++++++++------------------- obp/ope/meta.py | 24 +++++++++---------- obp/ope/regression_model.py | 4 ++-- obp/policy/base.py | 22 +++++++++--------- obp/policy/contextfree.py | 42 ++++++++++++++++----------------- obp/policy/linear.py | 26 ++++++++++----------- obp/policy/logistic.py | 34 +++++++++++++-------------- obp/policy/offline.py | 2 +- obp/utils.py | 10 ++++---- 12 files changed, 134 insertions(+), 132 deletions(-) diff --git a/obp/dataset/multiclass.py b/obp/dataset/multiclass.py index 1b50a070..5f2a2050 100644 --- a/obp/dataset/multiclass.py +++ b/obp/dataset/multiclass.py @@ -24,16 +24,16 @@ class MultiClassToBanditReduction(BaseSyntheticBanditDataset): A machine learning classifier such as logistic regression is used to construct behavior and evaluation policies as follows. 1. Split the original data into training (:math:`\\mathcal{D}_{\\mathrm{tr}}`) and evaluation (:math:`\\mathcal{D}_{\\mathrm{ev}}`) sets. - 2. Train classifiers on :math:`\\mathcal{D}_{\\mathrm{tr}}` and regard them as base deterministic policies :math:`\\pi_{\\mathrm{det},b}` and :math:`\\pi_{\\mathrm{det},e}`. - 3. Construct behavior (:math:`\\pi_{b}`) and evaluation (:math:`\\pi_{e}`) policies based on :math:`\\pi_{\\mathrm{det}}` as + 2. Train classifiers on :math:`\\mathcal{D}_{\\mathrm{tr}}` and obtain base deterministic policies :math:`\\pi_{\\mathrm{det},b}` and :math:`\\pi_{\\mathrm{det},e}`. + 3. Construct behavior (:math:`\\pi_{b}`) and evaluation (:math:`\\pi_{e}`) policies based on :math:`\\pi_{\\mathrm{det},b}` and :math:`\\pi_{\\mathrm{det},e}` as .. math:: - \\pi_b (a | x) := \\alpha_b \\pi_{\\mathrm{det},b} (a|x) + (1.0 - \\alpha_b) \\pi_{u} (a|x) + \\pi_b (a | x) := \\alpha_b \\cdot \\pi_{\\mathrm{det},b} (a|x) + (1.0 - \\alpha_b) \\cdot \\pi_{u} (a|x) .. math:: - \\pi_e (a | x) := \\alpha_e \\pi_{\\mathrm{det},e} (a|x) + (1.0 - \\alpha_e) \\pi_{u} (a|x) + \\pi_e (a | x) := \\alpha_e \\cdot \\pi_{\\mathrm{det},e} (a|x) + (1.0 - \\alpha_e) \\cdot \\pi_{u} (a|x) where :math:`\\pi_{u}` is a uniform random policy and :math:`\\alpha_b` and :math:`\\alpha_e` are set by the user. @@ -60,11 +60,11 @@ class MultiClassToBanditReduction(BaseSyntheticBanditDataset): base_classifier_b: ClassifierMixin Machine learning classifier used to construct a behavior policy. - alpha_b: float, default: 0.9 + alpha_b: float, default=0.9 Ration of a uniform random policy when constructing a **behavior** policy. Must be in the [0, 1) interval to make the behavior policy a stochastic one. - dataset_name: str, default: None + dataset_name: str, default=None Name of the dataset. Examples @@ -187,7 +187,7 @@ def split_train_eval( If float, should be between 0.0 and 1.0 and represent the proportion of the dataset to include in the evaluation split. If int, represents the absolute number of test samples. - random_state: int, default: None + random_state: int, default=None Controls the random seed in train-evaluation split. """ @@ -213,12 +213,12 @@ def obtain_batch_bandit_feedback( Please call `self.split_train_eval()` before calling this method. Parameters - ---------- + ----------- eval_size: float or int, default=0.25 If float, should be between 0.0 and 1.0 and represent the proportion of the dataset to include in the test split. If int, represents the absolute number of test samples. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. Returns @@ -261,10 +261,12 @@ def obtain_action_dist_by_eval_policy( ) -> np.ndarray: """Obtain action choice probabilities by an evaluation policy. - base_classifier_e: ClassifierMixin, default: None + Parameters + ----------- + base_classifier_e: ClassifierMixin, default=None Machine learning classifier used to construct a behavior policy. - alpha_e: float, default: 1.0 + alpha_e: float, default=1.0 Ration of a uniform random policy when constructing an **evaluation** policy. Must be in the [0, 1] interval (evaluation policy can be deterministic). diff --git a/obp/dataset/real.py b/obp/dataset/real.py index 861ef25d..605622e8 100644 --- a/obp/dataset/real.py +++ b/obp/dataset/real.py @@ -32,10 +32,10 @@ class OpenBanditDataset(BaseRealBanditDataset): campaign: str One of the three possible campaigns considered in ZOZOTOWN, "all", "men", and "women". - data_path: Path, default: Path('./obd') + data_path: Path, default=Path('./obd') Path that stores Open Bandit Dataset. - dataset_name: str, default: 'obd' + dataset_name: str, default='obd' Name of the dataset. References @@ -109,13 +109,13 @@ def calc_on_policy_policy_value_estimate( campaign: str One of the three possible campaigns considered in ZOZOTOWN (i.e., "all", "men", and "women"). - data_path: Path, default: Path('./obd') + data_path: Path, default=Path('./obd') Path that stores Open Bandit Dataset. test_size: float, default=0.3 If float, should be between 0.0 and 1.0 and represent the proportion of the dataset to include in the test split. - is_timeseries_split: bool, default: False + is_timeseries_split: bool, default=False If true, split the original logged badnit feedback data by time series. Returns @@ -178,7 +178,7 @@ def obtain_batch_bandit_feedback( test_size: float, default=0.3 If float, should be between 0.0 and 1.0 and represent the proportion of the dataset to include in the test split. - is_timeseries_split: bool, default: False + is_timeseries_split: bool, default=False If true, split the original logged badnit feedback data by time series. Returns @@ -233,10 +233,10 @@ def sample_bootstrap_bandit_feedback( test_size: float, default=0.3 If float, should be between 0.0 and 1.0 and represent the proportion of the dataset to include in the test split. - is_timeseries_split: bool, default: False + is_timeseries_split: bool, default=False If true, split the original logged badnit feedback data by time series. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling logged bandit dataset. Returns diff --git a/obp/dataset/synthetic.py b/obp/dataset/synthetic.py index 5e748acd..db4a9628 100644 --- a/obp/dataset/synthetic.py +++ b/obp/dataset/synthetic.py @@ -31,29 +31,29 @@ class SyntheticBanditDataset(BaseSyntheticBanditDataset): n_actions: int Number of actions. - dim_context: int, default: 1 + dim_context: int, default=1 Number of dimensions of context vectors. - reward_type: str, default: 'binary' + reward_type: str, default='binary' Type of reward variable, must be either 'binary' or 'continuous'. When 'binary' is given, rewards are sampled from the Bernoulli distribution. When 'continuous' is given, rewards are sampled from the truncated Normal distribution with `scale=1`. - reward_function: Callable[[np.ndarray, np.ndarray], np.ndarray]], default: None + reward_function: Callable[[np.ndarray, np.ndarray], np.ndarray]], default=None Function generating expected reward with context and action context vectors, i.e., :math:`\\mu: \\mathcal{X} \\times \\mathcal{A} \\rightarrow \\mathbb{R}`. If None is set, context **independent** expected reward for each action will be sampled from the uniform distribution automatically. - behavior_policy_function: Callable[[np.ndarray, np.ndarray], np.ndarray], default: None + behavior_policy_function: Callable[[np.ndarray, np.ndarray], np.ndarray], default=None Function generating probability distribution over action space, i.e., :math:`\\pi: \\mathcal{X} \\rightarrow \\Delta(\\mathcal{A})`. If None is set, context **independent** uniform distribution will be used (uniform random behavior policy). - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling synthetic bandit dataset. - dataset_name: str, default: 'synthetic_bandit_dataset' + dataset_name: str, default='synthetic_bandit_dataset' Name of the dataset. Examples @@ -252,7 +252,7 @@ def logistic_reward_function( action_context: array-like, shape (n_actions, dim_action_context) Vector representation for each action. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling dataset. Returns @@ -292,7 +292,7 @@ def linear_reward_function( action_context: array-like, shape (n_actions, dim_action_context) Vector representation for each action. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling dataset. Returns @@ -332,7 +332,7 @@ def linear_behavior_policy( action_context: array-like, shape (n_actions, dim_action_context) Vector representation for each action. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling dataset. Returns diff --git a/obp/ope/estimators.py b/obp/ope/estimators.py index 6fd87544..45d29d1f 100644 --- a/obp/ope/estimators.py +++ b/obp/ope/estimators.py @@ -49,7 +49,7 @@ class ReplayMethod(BaseOffPolicyEstimator): Parameters ---------- - estimator_name: str, default: 'rm'. + estimator_name: str, default='rm'. Name of off-policy estimator. References @@ -151,13 +151,13 @@ def estimate_interval( position: array-like, shape (n_rounds,) Positions of each round in the given logged bandit feedback. - alpha: float, default: 0.05 + alpha: float, default=0.05 P-value. - n_bootstrap_samples: int, default: 10000 + n_bootstrap_samples: int, default=10000 Number of resampling performed in the bootstrap procedure. - random_state: int, default: None + random_state: int, default=None Controls the random seed in bootstrap sampling. Returns @@ -197,7 +197,7 @@ class InverseProbabilityWeighting(BaseOffPolicyEstimator): Parameters ------------ - estimator_name: str, default: 'ipw'. + estimator_name: str, default='ipw'. Name of off-policy estimator. References @@ -320,13 +320,13 @@ def estimate_interval( Distribution over actions or the action choice probabilities by the evaluation policy (can be deterministic), i.e., :math:`\\pi_e(a|x)`. - alpha: float, default: 0.05 + alpha: float, default=0.05 P-value. - n_bootstrap_samples: int, default: 10000 + n_bootstrap_samples: int, default=10000 Number of resampling performed in the bootstrap procedure. - random_state: int, default: None + random_state: int, default=None Controls the random seed in bootstrap sampling. Returns @@ -372,7 +372,7 @@ class SelfNormalizedInverseProbabilityWeighting(InverseProbabilityWeighting): Parameters ---------- - estimator_name: str, default: 'snipw'. + estimator_name: str, default='snipw'. Name of off-policy estimator. References @@ -451,7 +451,7 @@ class DirectMethod(BaseOffPolicyEstimator): Parameters ---------- - estimator_name: str, default: 'dm'. + estimator_name: str, default='dm'. Name of off-policy estimator. References @@ -554,13 +554,13 @@ def estimate_interval( estimated_rewards_by_reg_model: array-like, shape (n_rounds, n_actions, len_list) Estimated rewards for each round, action, and position by regression model, i.e., :math:`\\hat{q}(x_t,a_t)`. - alpha: float, default: 0.05 + alpha: float, default=0.05 P-value. - n_bootstrap_samples: int, default: 10000 + n_bootstrap_samples: int, default=10000 Number of resampling performed in the bootstrap procedure. - random_state: int, default: None + random_state: int, default=None Controls the random seed in bootstrap sampling. Returns @@ -611,7 +611,7 @@ class DoublyRobust(InverseProbabilityWeighting): Parameters ---------- - estimator_name: str, default: 'dr'. + estimator_name: str, default='dr'. Name of off-policy estimator. References @@ -760,13 +760,13 @@ def estimate_interval( estimated_rewards_by_reg_model: array-like, shape (n_rounds, n_actions, len_list) Estimated rewards for each round, action, and position by regression model, i.e., :math:`\\hat{q}(x_t,a_t)`. - alpha: float, default: 0.05 + alpha: float, default=0.05 P-value. - n_bootstrap_samples: int, default: 10000 + n_bootstrap_samples: int, default=10000 Number of resampling performed in the bootstrap procedure. - random_state: int, default: None + random_state: int, default=None Controls the random seed in bootstrap sampling. Returns @@ -815,7 +815,7 @@ class SelfNormalizedDoublyRobust(DoublyRobust): Parameters ---------- - estimator_name: str, default: 'sndr'. + estimator_name: str, default='sndr'. Name of off-policy estimator. References @@ -906,11 +906,11 @@ class SwitchInverseProbabilityWeighting(DoublyRobust): Parameters ---------- - tau: float, default: 1 + tau: float, default=1 Switching hyperparameter. When importance weight is larger than this parameter, the DM estimator is applied, otherwise the IPW estimator is applied. This hyperparameter should be larger than 1., otherwise it is meaningless. - estimator_name: str, default: 'switch-ipw'. + estimator_name: str, default='switch-ipw'. Name of off-policy estimator. References @@ -1007,11 +1007,11 @@ class SwitchDoublyRobust(DoublyRobust): Parameters ---------- - tau: float, default: 1 + tau: float, default=1 Switching hyperparameter. When importance weight is larger than this parameter, the DM estimator is applied, otherwise the DR estimator is applied. This hyperparameter should be larger than 0., otherwise it is meaningless. - estimator_name: str, default: 'switch-dr'. + estimator_name: str, default='switch-dr'. Name of off-policy estimator. References @@ -1127,7 +1127,7 @@ class DoublyRobustWithShrinkage(DoublyRobust): lambda_: float Shrinkage hyperparameter. This hyperparameter should be larger than 0., otherwise it is meaningless. - estimator_name: str, default: 'dr-os'. + estimator_name: str, default='dr-os'. Name of off-policy estimator. References diff --git a/obp/ope/meta.py b/obp/ope/meta.py index 4bf13393..ce09e523 100644 --- a/obp/ope/meta.py +++ b/obp/ope/meta.py @@ -161,13 +161,13 @@ def estimate_intervals( Estimated expected rewards for the given logged bandit feedback at each item and position by regression model. When it is not given, model-dependent estimators such as DM and DR cannot be used. - alpha: float, default: 0.05 + alpha: float, default=0.05 P-value. - n_bootstrap_samples: int, default: 100 + n_bootstrap_samples: int, default=100 Number of resampling performed in the bootstrap procedure. - random_state: int, default: None + random_state: int, default=None Controls the random seed in bootstrap sampling. Returns @@ -218,13 +218,13 @@ def summarize_off_policy_estimates( Estimated expected rewards for the given logged bandit feedback at each item and position by regression model. When it is not given, model-dependent estimators such as DM and DR cannot be used. - alpha: float, default: 0.05 + alpha: float, default=0.05 P-value. - n_bootstrap_samples: int, default: 100 + n_bootstrap_samples: int, default=100 Number of resampling performed in the bootstrap procedure. - random_state: int, default: None + random_state: int, default=None Controls the random seed in bootstrap sampling. Returns @@ -277,24 +277,24 @@ def visualize_off_policy_estimates( Estimated expected rewards for the given logged bandit feedback at each item and position by regression model. When it is not given, model-dependent estimators such as DM and DR cannot be used. - alpha: float, default: 0.05 + alpha: float, default=0.05 P-value. - n_bootstrap_samples: int, default: 100 + n_bootstrap_samples: int, default=100 Number of resampling performed in the bootstrap procedure. - random_state: int, default: None + random_state: int, default=None Controls the random seed in bootstrap sampling. - is_relative: bool, default: False, + is_relative: bool, default=False, If True, the method visualizes the estimated policy values of evaluation policy relative to the ground-truth policy value of behavior policy. - fig_dir: Path, default: None + fig_dir: Path, default=None Path to store the bar figure. If 'None' is given, the figure will not be saved. - fig_name: str, default: "estimated_policy_value.png" + fig_name: str, default="estimated_policy_value.png" Name of the bar figure. """ diff --git a/obp/ope/regression_model.py b/obp/ope/regression_model.py index 6cd501fc..d429752e 100644 --- a/obp/ope/regression_model.py +++ b/obp/ope/regression_model.py @@ -28,7 +28,7 @@ class RegressionModel(BaseEstimator): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. @@ -101,7 +101,7 @@ def fit( reward: array-like, shape (n_rounds,) Observed rewards (or outcome) in each round, i.e., :math:`r_t`. - pscore: Optional[np.ndarray], default: None + pscore: Optional[np.ndarray], default=None Propensity scores, the action choice probabilities by behavior policy, in the training logged bandit feedback. diff --git a/obp/policy/base.py b/obp/policy/base.py index d2a428a0..c0f363ae 100644 --- a/obp/policy/base.py +++ b/obp/policy/base.py @@ -22,14 +22,14 @@ class BaseContextFreePolicy(metaclass=ABCMeta): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. - batch_size: int, default: 1 + batch_size: int, default=1 Number of samples used in a batch parameter update. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. """ @@ -95,20 +95,20 @@ class BaseContextualPolicy(metaclass=ABCMeta): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. - batch_size: int, default: 1 + batch_size: int, default=1 Number of samples used in a batch parameter update. - alpha_: float, default: 1. + alpha_: float, default=1. Prior parameter for the online logistic regression. - lambda_: float, default: 1. + lambda_: float, default=1. Regularization hyperparameter for the online logistic regression. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. """ @@ -180,7 +180,7 @@ class BaseOffPolicyLearner(metaclass=ABCMeta): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. @@ -234,7 +234,7 @@ def _create_train_data_for_opl( reward: array-like, shape (n_actions,) Observed rewards (or outcome) in each round, i.e., :math:`r_t`. - pscore: array-like, shape (n_actions,), default: None + pscore: array-like, shape (n_actions,), default=None Propensity scores or the action choice probabilities by behavior policy, i.e., :math:`\\pi_b(a_t|x_t)`. Returns @@ -266,7 +266,7 @@ def fit( reward: array-like, shape (n_rounds,) Observed rewards (or outcome) in each round, i.e., :math:`r_t`. - pscore: array-like, shape (n_rounds,), default: None + pscore: array-like, shape (n_rounds,), default=None Propensity scores or the action choice probabilities by behavior policy, i.e., :math:`\\pi_b(a_t|x_t)`. position: array-like, shape (n_rounds,), default=None diff --git a/obp/policy/contextfree.py b/obp/policy/contextfree.py index 0b893d17..e8072d03 100644 --- a/obp/policy/contextfree.py +++ b/obp/policy/contextfree.py @@ -30,20 +30,20 @@ class EpsilonGreedy(BaseContextFreePolicy): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. - batch_size: int, default: 1 + batch_size: int, default=1 Number of samples used in a batch parameter update. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. - epsilon: float, default: 1. + epsilon: float, default=1. Exploration hyperparameter that must take value in the range of [0., 1.]. - policy_name: str, default: f'egreedy_{epsilon}'. + policy_name: str, default=f'egreedy_{epsilon}'. Name of bandit policy. """ @@ -105,20 +105,20 @@ class Random(EpsilonGreedy): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. - batch_size: int, default: 1 + batch_size: int, default=1 Number of samples used in a batch parameter update. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. - epsilon: float, default: 1. + epsilon: float, default=1. Exploration hyperparameter that must take value in the range of [0., 1.]. - policy_name: str, default: 'random'. + policy_name: str, default='random'. Name of bandit policy. """ @@ -132,7 +132,7 @@ def compute_batch_action_dist( Parameters ---------- - n_rounds: int, default: 1 + n_rounds: int, default=1 Number of rounds in the distribution over actions. (the size of the first axis of `action_dist`) @@ -157,30 +157,30 @@ class BernoulliTS(BaseContextFreePolicy): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. - batch_size: int, default: 1 + batch_size: int, default=1 Number of samples used in a batch parameter update. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. - alpha: array-like, shape (n_actions, ), default: None + alpha: array-like, shape (n_actions, ), default=None Prior parameter vector for Beta distributions. - beta: array-like, shape (n_actions, ), default: None + beta: array-like, shape (n_actions, ), default=None Prior parameter vector for Beta distributions. - is_zozotown_prior: bool, default: False + is_zozotown_prior: bool, default=False Whether to use hyperparameters for the beta distribution used at the start of the data collection period in ZOZOTOWN. - campaign: str, default: None + campaign: str, default=None One of the three possible campaigns considered in ZOZOTOWN, "all", "men", and "women". - policy_name: str, default: 'bts' + policy_name: str, default='bts' Name of bandit policy. """ @@ -245,11 +245,11 @@ def compute_batch_action_dist( Parameters ---------- - n_rounds: int, default: 1 + n_rounds: int, default=1 Number of rounds in the distribution over actions. (the size of the first axis of `action_dist`) - n_sim: int, default: 100000 + n_sim: int, default=100000 Number of simulations in the Monte Carlo simulation to compute the distribution over actions. Returns diff --git a/obp/policy/linear.py b/obp/policy/linear.py index 63f9a6dd..d63b25e9 100644 --- a/obp/policy/linear.py +++ b/obp/policy/linear.py @@ -21,20 +21,20 @@ class LinEpsilonGreedy(BaseContextualPolicy): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. - batch_size: int, default: 1 + batch_size: int, default=1 Number of samples used in a batch parameter update. - n_trial: int, default: 0 + n_trial: int, default=0 Current number of trials in a bandit simulation. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. - epsilon: float, default: 0. + epsilon: float, default=0. Exploration hyperparameter that must take value in the range of [0., 1.]. References @@ -140,17 +140,17 @@ class LinUCB(BaseContextualPolicy): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. - batch_size: int, default: 1 + batch_size: int, default=1 Number of samples used in a batch parameter update. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. - epsilon: float, default: 0. + epsilon: float, default=0. Exploration hyperparameter that must take value in the range of [0., 1.]. References @@ -257,17 +257,17 @@ class LinTS(BaseContextualPolicy): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. - batch_size: int, default: 1 + batch_size: int, default=1 Number of samples used in a batch parameter update. - alpha_: float, default: 1. + alpha_: float, default=1. Prior parameter for the online logistic regression. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. """ diff --git a/obp/policy/logistic.py b/obp/policy/logistic.py index a2d39604..c8945791 100644 --- a/obp/policy/logistic.py +++ b/obp/policy/logistic.py @@ -25,23 +25,23 @@ class LogisticEpsilonGreedy(BaseContextualPolicy): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. - batch_size: int, default: 1 + batch_size: int, default=1 Number of samples used in a batch parameter update. - alpha_: float, default: 1. + alpha_: float, default=1. Prior parameter for the online logistic regression. - lambda_: float, default: 1. + lambda_: float, default=1. Regularization hyperparameter for the online logistic regression. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. - epsilon: float, default: 0. + epsilon: float, default=0. Exploration hyperparameter that must take value in the range of [0., 1.]. """ @@ -131,23 +131,23 @@ class LogisticUCB(BaseContextualPolicy): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. - batch_size: int, default: 1 + batch_size: int, default=1 Number of samples used in a batch parameter update. - alpha_: float, default: 1. + alpha_: float, default=1. Prior parameter for the online logistic regression. - lambda_: float, default: 1. + lambda_: float, default=1. Regularization hyperparameter for the online logistic regression. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. - epsilon: float, default: 0. + epsilon: float, default=0. Exploration hyperparameter that must take value in the range of [0., 1.]. References @@ -244,20 +244,20 @@ class LogisticTS(BaseContextualPolicy): n_actions: int Number of actions. - len_list: int, default: 1 + len_list: int, default=1 Length of a list of recommended actions in each impression. When Open Bandit Dataset is used, 3 should be set. - batch_size: int, default: 1 + batch_size: int, default=1 Number of samples used in a batch parameter update. - alpha_: float, default: 1. + alpha_: float, default=1. Prior parameter for the online logistic regression. - lambda_: float, default: 1. + lambda_: float, default=1. Regularization hyperparameter for the online logistic regression. - random_state: int, default: None + random_state: int, default=None Controls the random seed in sampling actions. References diff --git a/obp/policy/offline.py b/obp/policy/offline.py index 7684e7b8..d7944268 100644 --- a/obp/policy/offline.py +++ b/obp/policy/offline.py @@ -51,7 +51,7 @@ def _create_train_data_for_opl( reward: array-like, shape (n_rounds,) Observed rewards (or outcome) in each round, i.e., :math:`r_t`. - pscore: array-like, shape (n_rounds,), default: None + pscore: array-like, shape (n_rounds,), default=None Propensity scores, the probability of selecting each action by behavior policy, in the given logged bandit feedback. diff --git a/obp/utils.py b/obp/utils.py index 4ae9473d..04fe2052 100644 --- a/obp/utils.py +++ b/obp/utils.py @@ -24,13 +24,13 @@ def estimate_confidence_interval_by_bootstrap( samples: array-like Empirical observed samples to be used to estimate cumulative distribution function. - alpha: float, default: 0.05 + alpha: float, default=0.05 P-value. - n_bootstrap_samples: int, default: 10000 + n_bootstrap_samples: int, default=10000 Number of resampling performed in the bootstrap procedure. - random_state: int, default: None + random_state: int, default=None Controls the random seed in bootstrap sampling. Returns @@ -179,10 +179,10 @@ def check_bandit_feedback_inputs( reward: array-like, shape (n_rounds,) Observed rewards (or outcome) in each round, i.e., :math:`r_t`. - position: array-like, shape (n_rounds,), default: None + position: array-like, shape (n_rounds,), default=None Positions of each round in the given logged bandit feedback. - pscore: array-like, shape (n_rounds,), default: None + pscore: array-like, shape (n_rounds,), default=None Propensity scores, the probability of selecting each action by behavior policy, in the given logged bandit feedback. From f09f2180b1b52f109848b524b5dfeff0732e6446 Mon Sep 17 00:00:00 2001 From: usaito Date: Sun, 8 Nov 2020 01:45:03 +0900 Subject: [PATCH 4/4] update version --- obp/version.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/obp/version.py b/obp/version.py index 260c070a..f9aa3e11 100644 --- a/obp/version.py +++ b/obp/version.py @@ -1 +1 @@ -__version__ = "0.3.1" +__version__ = "0.3.2"