add source

.gitignore

@@ -0,0 +1,8 @@
+__pycache__
+.ipynb_checkpoints
+build
+*.egg-info
+dist
+htmlcov
+.coverage
+.mypy*

Makefile

@@ -0,0 +1,27 @@
+TESTS_DIR=tests
+
+format:
+	yapf -irp .
+
+dead-code-check:
+	vulture src --min-confidence 70
+
+test:
+	python -m unittest discover ${TESTS_DIR}
+
+type-check:
+	mypy src --ignore-missing-imports
+
+coverage:
+	coverage run --source=src -m unittest discover ${TESTS_DIR}
+	coverage report -m
+	coverage html
+
+build:
+	rm -rf dist
+	python -m pip install --upgrade build
+	python -m build
+	twine check dist/*
+
+push:
+	twine upload dist/*

notebooks/figure_2.ipynb

@@ -0,0 +1,244 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "2a5d14b3-49cd-4ab3-990b-4e1951566381",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from datetime import datetime"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "a2a441d6-4985-482e-9a6e-e402cee100e8",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import gorillacompression as gc"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "744bd485-b1b8-4b66-adfd-f765ed145e43",
+   "metadata": {},
+   "source": [
+    "# Figure 2"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "f6906a26-530e-441e-a183-13e0359904f5",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "timestamps_str = ['March 24, 2015 02:01:02', 'March 24, 2015 02:02:02', 'March 24, 2015 02:03:02']\n",
+    "timestamps_dt = list(map(lambda x: datetime.strptime(x, '%B %d, %Y %X'), timestamps_str))\n",
+    "timestamps_int = [int(ts.timestamp()) for ts in timestamps_dt]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "fa8c693a-bca0-4ec8-82ae-9ee60dc9cc11",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[1427158862, 1427158922, 1427158982]"
+      ]
+     },
+     "execution_count": 4,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "timestamps_int"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "a81347f0-6828-40ab-a554-9bb7fdbd764d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "values = [12.0, 12.0, 24.0]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "512ef064-d587-4d41-a1ef-14f60443690c",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[12.0, 12.0, 24.0]"
+      ]
+     },
+     "execution_count": 6,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "values"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "c7a301c3-e967-42a6-8da1-429b7548fcc6",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# content = gc.PairsEncoder.encode_all(list(zip(timestamps_int, values)))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "7b05b2cb-e4b3-4641-afdf-0425d472b33c",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "pairs_encoder = gc.PairsEncoder()\n",
+    "\n",
+    "for ts, v in zip(timestamps_int, values):\n",
+    "    pairs_encoder.encode_next(ts, v)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "id": "e74710f7-82a5-4756-b6e8-b1de18f3523d",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "bitarray('1010101000100001011011101001110010000000010100000000000000000000000000000000000000000000000000000011010110000011')"
+      ]
+     },
+     "execution_count": 9,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "pairs_encoder.bit_array"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "7d304e17-4a18-43df-ac46-1683e081b7ae",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "content = pairs_encoder.get_encoded()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "id": "1cbb6a03-b147-4244-b236-d187dadde534",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "{'encoded': b'\\xaa!n\\x9c\\x80P\\x00\\x00\\x00\\x00\\x00\\x005\\x83',\n",
+       " 'nb_pairs': 3,\n",
+       " 'float_format': 'f64'}"
+      ]
+     },
+     "execution_count": 11,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "content"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "id": "506398d0-73d1-4360-84f2-823b7711a47e",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "14"
+      ]
+     },
+     "execution_count": 12,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "len(content['encoded'])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "id": "966396b9-4b7f-4c91-9dea-eb2aa47a2a74",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[(1427158862, 12.0), (1427158922, 12.0), (1427158982, 24.0)]"
+      ]
+     },
+     "execution_count": 13,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "gc.PairsDecoder.decode_all(content)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "c374aef9-4bdc-43f6-877e-9fcdb567d106",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "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": "3.8.10"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}