Created
June 18, 2021 18:04
-
-
Save alyferryhalo/390219047b02a025334ae5c0d0583a6e to your computer and use it in GitHub Desktop.
time_series.ipynb
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| { | |
| "nbformat": 4, | |
| "nbformat_minor": 0, | |
| "metadata": { | |
| "colab": { | |
| "name": "time_series.ipynb", | |
| "provenance": [], | |
| "collapsed_sections": [], | |
| "authorship_tag": "ABX9TyPq4EC8UCBOPZagVWcJbk0I", | |
| "include_colab_link": true | |
| }, | |
| "kernelspec": { | |
| "name": "python3", | |
| "display_name": "Python 3" | |
| }, | |
| "language_info": { | |
| "name": "python" | |
| } | |
| }, | |
| "cells": [ | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "view-in-github", | |
| "colab_type": "text" | |
| }, | |
| "source": [ | |
| "<a href=\"https://colab.research.google.com/gist/alyferryhalo/390219047b02a025334ae5c0d0583a6e/time_series.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "ISGpmOjhB-97" | |
| }, | |
| "source": [ | |
| "# Анализ временных рядов\n", | |
| "\n", | |
| "Данный ноутбук посвящён анализу временных рядов на языке программирования Python, в частности, с использованием библиотеки pandas. Со списком литературы и иными ресурсами можно ознакомиться в конце работы.\n", | |
| "\n", | |
| "" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "IXFEk1UkrIdL" | |
| }, | |
| "source": [ | |
| "## Введение\n", | |
| "\n", | |
| "Запросы на прогнозирование во времени, например, числа продаж, уровня заболеваемости, концентрации лекарственного препарата, уровня сахара в крови у пациента, достаточно часты и актуальны в повседневной жизни. Они зависят от множества факторов, но могут быть спрогнозированы с определённой точностью — в зависимости от полноты исходных данных и качестве моделей, разумеется.\n", | |
| "\n", | |
| "Именно решением подобных задач мы и займёмся.\n", | |
| "\n", | |
| "В соответствии с определением,\n", | |
| "\n", | |
| "> **Временной ряд** (или ряд динамики) — собранный в разные моменты времени статистический материал о значении каких-либо параметров (в простейшем случае одного) исследуемого процесса. Каждая единица статистического материала называется измерением или отсчётом, также допустимо называть его уровнем на указанный с ним момент времени. Во временном ряде для каждого отсчёта должно быть указано время измерения или номер измерения по порядку. Временной ряд существенно отличается от простой выборки данных, так как при анализе учитывается взаимосвязь измерений со временем, а не только статистическое разнообразие и статистические характеристики выборки.\n", | |
| "\n", | |
| "Временные ряды, как правило, возникают в результате измерения некоторого показателя. Это могут быть как показатели (характеристики) технических систем, так и показатели природных, социальных, экономических и других систем (например, погодные данные). Типичным примером временного ряда можно назвать биржевой курс, при анализе которого пытаются определить основное направление развития (тенденцию или тренд). \n" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "DeVhoS4ovqCp" | |
| }, | |
| "source": [ | |
| "### Базовые понятия" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "VmecvQB_uJDT" | |
| }, | |
| "source": [ | |
| "Прежде всего сделаем некоторые полезные импорты:" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "rao7Xu5St3KH" | |
| }, | |
| "source": [ | |
| "import warnings\n", | |
| "warnings.filterwarnings('ignore')\n", | |
| "\n", | |
| "from itertools import product\n", | |
| "from tqdm import tqdm_notebook\n", | |
| "\n", | |
| "import numpy as np\n", | |
| "import pandas as pd\n", | |
| "import matplotlib.pyplot as plt\n", | |
| "import seaborn as sns\n", | |
| "\n", | |
| "from scipy.optimize import minimize\n", | |
| "import scipy.stats as scs\n", | |
| "\n", | |
| "import statsmodels.formula.api as smf\n", | |
| "import statsmodels.tsa.api as smt\n", | |
| "import statsmodels.api as sm\n", | |
| "\n", | |
| "from dateutil.relativedelta import relativedelta\n", | |
| "\n", | |
| "%matplotlib inline" | |
| ], | |
| "execution_count": null, | |
| "outputs": [] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "Lgg5SGZUZa8I" | |
| }, | |
| "source": [ | |
| "Что касается библиотек:\n", | |
| "\n", | |
| "* numpy\n", | |
| "* pandas\n", | |
| "* matplotlib\n", | |
| "* seaborn\n", | |
| "* scipy\n", | |
| "* statsmodels\n", | |
| "* dateutil\n" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "pxfOL912u7HC" | |
| }, | |
| "source": [ | |
| "А в качестве примера импортируем датафрейм:" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/", | |
| "height": 206 | |
| }, | |
| "id": "TLjW5G7qupYk", | |
| "outputId": "81ae62f4-59e6-48c4-c6fb-66f74c7e4e00" | |
| }, | |
| "source": [ | |
| "df = pd.read_csv('https://raw.githubusercontent.com/selva86/datasets/master/a10.csv', parse_dates=['date'])\n", | |
| "df.head()" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "execute_result", | |
| "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>date</th>\n", | |
| " <th>value</th>\n", | |
| " </tr>\n", | |
| " </thead>\n", | |
| " <tbody>\n", | |
| " <tr>\n", | |
| " <th>0</th>\n", | |
| " <td>1991-07-01</td>\n", | |
| " <td>3.526591</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>1</th>\n", | |
| " <td>1991-08-01</td>\n", | |
| " <td>3.180891</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>2</th>\n", | |
| " <td>1991-09-01</td>\n", | |
| " <td>3.252221</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>3</th>\n", | |
| " <td>1991-10-01</td>\n", | |
| " <td>3.611003</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>4</th>\n", | |
| " <td>1991-11-01</td>\n", | |
| " <td>3.565869</td>\n", | |
| " </tr>\n", | |
| " </tbody>\n", | |
| "</table>\n", | |
| "</div>" | |
| ], | |
| "text/plain": [ | |
| " date value\n", | |
| "0 1991-07-01 3.526591\n", | |
| "1 1991-08-01 3.180891\n", | |
| "2 1991-09-01 3.252221\n", | |
| "3 1991-10-01 3.611003\n", | |
| "4 1991-11-01 3.565869" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 2 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/", | |
| "height": 444 | |
| }, | |
| "id": "MJMmBCTPvGyb", | |
| "outputId": "210ae7f0-171e-415c-ab05-1753da008e71" | |
| }, | |
| "source": [ | |
| "plt.figure(figsize=(15, 7))\n", | |
| "plt.plot(df.value)\n", | |
| "plt.title('Australian Drug Sales')\n", | |
| "plt.grid(True)\n", | |
| "plt.show()" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "display_data", | |
| "data": { | |
| "image/png": "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\n", | |
| "text/plain": [ | |
| "<Figure size 1080x504 with 1 Axes>" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [], | |
| "needs_background": "light" | |
| } | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "NTs5ur7irP7J" | |
| }, | |
| "source": [ | |
| "### Метрики качества" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "hyT6-9EgrS9a" | |
| }, | |
| "source": [ | |
| "Коэффициент детерминации (R^2 — R-квадрат) — это доля дисперсии зависимой переменной, объясняемая рассматриваемой моделью зависимости, то есть объясняющими переменными. Более точно — это единица минус доля необъяснённой дисперсии (дисперсии случайной ошибки модели, или условной по факторам дисперсии зависимой переменной) в дисперсии зависимой переменной. Его рассматривают как универсальную меру зависимости одной случайной величины от множества других. \n", | |
| "\n", | |
| "Принимает значения в диапазоне (−∞,1]\n", | |
| "\n", | |
| "Можно найти в sklearn:\n", | |
| "```\n", | |
| "sklearn.metrics.r2_score\n", | |
| "```\n", | |
| "\n" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "cec6ORs8R-7j" | |
| }, | |
| "source": [ | |
| "**Средняя абсолютная ошибка (MAE)** — это мера ошибок между парными наблюдениями, выражающими одно и то же явление. Примеры Y по сравнению с X включают сравнения прогнозируемого и наблюдаемого, последующего времени и начального времени, а также один метод измерения по сравнению с альтернативным методом измерения.\n", | |
| "\n", | |
| "Принимает значения в диапазоне [0,+∞)\n", | |
| "\n", | |
| "```\n", | |
| "sklearn.metrics.mean_absolute_error\n", | |
| "```\n", | |
| "\n", | |
| "" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "q8YceGqESAOJ" | |
| }, | |
| "source": [ | |
| "**Среднеквадратичная ошибка** (MSE) измеряет среднее значение квадратов ошибок, то есть среднеквадратичную разницумежду расчетными и фактическими значениями.\n", | |
| "\n", | |
| "Принимает значения в диапазоне [0,+∞)\n", | |
| "\n", | |
| "```\n", | |
| "sklearn.metrics.mean_squared_error\n", | |
| "```\n", | |
| "\n", | |
| "" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "SXmnBRCerpyT" | |
| }, | |
| "source": [ | |
| "## Список литературы и дополнительные материалы\n" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "XxFVFVJesB81" | |
| }, | |
| "source": [ | |
| "### Интернет-статьи, ноутбуки...\n", | |
| "\n", | |
| "1. [Kaggle: Time series analysis in Python](https://www.kaggle.com/kashnitsky/topic-9-part-1-time-series-analysis-in-python)\n", | |
| "2. [Time Series Analysis in Python – A Comprehensive Guide with Examples](https://www.machinelearningplus.com/time-series-analysis-python/)\n", | |
| "3. [Time Series Analysis in Python: An Introduction](https://towardsdatascience.com/time-series-analysis-in-python-an-introduction-70d5a5b1d52a)\n", | |
| "4. [Временной ряд](https://ru.wikipedia.org/wiki/%D0%92%D1%80%D0%B5%D0%BC%D0%B5%D0%BD%D0%BD%D0%BE%D0%B9_%D1%80%D1%8F%D0%B4)\n", | |
| "5. [Открытый курс машинного обучения. Тема 9. Анализ временных рядов с помощью Python](https://habr.com/ru/company/ods/blog/327242/)" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "zS6R5jQAsHPB" | |
| }, | |
| "source": [ | |
| "### Видео\n", | |
| "\n", | |
| "\n", | |
| "\n", | |
| "1. [Лекция 9. Временные ряды. Открытый курс OpenDataScience по машинному обучению](https://www.youtube.com/watch?v=nQjul-5_0_M)\n", | |
| "2. [Анализ временных рядов. Практика по машинному обучению.](https://www.youtube.com/watch?v=dpIz4qj_MMY)\n", | |
| "\n" | |
| ] | |
| } | |
| ] | |
| } |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment