{ "nbformat": 4, "nbformat_minor": 0, "metadata": { "colab": { "provenance": [] }, "kernelspec": { "name": "python3", "display_name": "Python 3" }, "language_info": { "name": "python" } }, "cells": [ { "cell_type": "markdown", "source": [ "# 📅 Работа с датой и временем в Pandas" ], "metadata": { "id": "EpIvo1sidsYX" } }, { "cell_type": "markdown", "source": [ "В Data Science время — это не просто строка. Это ось, по которой мы можем анализировать тренды, считать длительность событий и группировать данные.\n", "В этом ноутбуке разберем:\n", "1. Создание datetime объектов.\n", "2. Извлечение признаков (Feature Extraction).\n", "3. Математические операции со временем." ], "metadata": { "id": "WZ9keSr2dwf2" } }, { "cell_type": "code", "source": [ "import pandas as pd\n", "import numpy as np" ], "metadata": { "id": "Gq9CvdfGd05l" }, "execution_count": 1, "outputs": [] }, { "cell_type": "markdown", "source": [ "## Преобразование в datetime" ], "metadata": { "id": "5WgSw0HWd34F" } }, { "cell_type": "markdown", "source": [ "Часто данные приходят в виде строк. Pandas умеет автоматически или по заданному формату парсить их в тип `datetime64[ns]`" ], "metadata": { "id": "Mm3Q-XtAd7wN" } }, { "cell_type": "code", "source": [ "# Пример: логи входа игроков в Rublox\n", "logins = ['2026-11-08 14:30:00', '2026-11-08 15:45:00', '2026-11-09 10:00:00']\n", "\n", "# Преобразуем в datetime\n", "dt_series = pd.to_datetime(logins)\n", "print(dt_series)\n", "print(f\"Тип данных: {dt_series.dtype}\")" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "uEuwkYPVd_lL", "outputId": "d0dd72c4-ad12-444b-fdd3-0c2da7ae4281" }, "execution_count": 5, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "DatetimeIndex(['2026-11-08 14:30:00', '2026-11-08 15:45:00',\n", " '2026-11-09 10:00:00'],\n", " dtype='datetime64[ns]', freq=None)\n", "Тип данных: datetime64[ns]\n" ] } ] }, { "cell_type": "markdown", "source": [ "## Извлечение компонентов" ], "metadata": { "id": "iIsepuMleUiN" } }, { "cell_type": "markdown", "source": [ "Как только столбец стал типом `datetime`, мы можем использовать аксессор `.dt`, чтобы \"вытащить\" из него год, месяц, день недели и т.д. Это супер-полезно для аналитики!" ], "metadata": { "id": "cy5jIPiEeYfN" } }, { "cell_type": "code", "source": [ "df = pd.DataFrame({'login_time': dt_series})\n", "\n", "# Извлекаем признаки\n", "df['year'] = df['login_time'].dt.year\n", "df['month'] = df['login_time'].dt.month\n", "df['day'] = df['login_time'].dt.day\n", "df['hour'] = df['login_time'].dt.hour\n", "df['dayofweek'] = df['login_time'].dt.dayofweek # 0 = Понедельник, 6 = Воскресенье\n", "df['day_name'] = df['login_time'].dt.day_name() # Название дня\n", "\n", "df" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 143 }, "id": "-0KF54uxec49", "outputId": "be6a7f63-d8ba-4241-83f1-52116cb27fe7" }, "execution_count": 6, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ " login_time year month day hour dayofweek day_name\n", "0 2026-11-08 14:30:00 2026 11 8 14 6 Sunday\n", "1 2026-11-08 15:45:00 2026 11 8 15 6 Sunday\n", "2 2026-11-09 10:00:00 2026 11 9 10 0 Monday" ], "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", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
login_timeyearmonthdayhourdayofweekday_name
02026-11-08 14:30:002026118146Sunday
12026-11-08 15:45:002026118156Sunday
22026-11-09 10:00:002026119100Monday
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "df", "summary": "{\n \"name\": \"df\",\n \"rows\": 3,\n \"fields\": [\n {\n \"column\": \"login_time\",\n \"properties\": {\n \"dtype\": \"date\",\n \"min\": \"2026-11-08 14:30:00\",\n \"max\": \"2026-11-09 10:00:00\",\n \"num_unique_values\": 3,\n \"samples\": [\n \"2026-11-08 14:30:00\",\n \"2026-11-08 15:45:00\",\n \"2026-11-09 10:00:00\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"year\",\n \"properties\": {\n \"dtype\": \"int32\",\n \"num_unique_values\": 1,\n \"samples\": [\n 2026\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"month\",\n \"properties\": {\n \"dtype\": \"int32\",\n \"num_unique_values\": 1,\n \"samples\": [\n 11\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"day\",\n \"properties\": {\n \"dtype\": \"int32\",\n \"num_unique_values\": 2,\n \"samples\": [\n 9\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"hour\",\n \"properties\": {\n \"dtype\": \"int32\",\n \"num_unique_values\": 3,\n \"samples\": [\n 14\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"dayofweek\",\n \"properties\": {\n \"dtype\": \"int32\",\n \"num_unique_values\": 2,\n \"samples\": [\n 0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"day_name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 2,\n \"samples\": [\n \"Monday\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}" } }, "metadata": {}, "execution_count": 6 } ] }, { "cell_type": "markdown", "source": [ "## Математика времени Timedelta" ], "metadata": { "id": "pR7I5MoiefAU" } }, { "cell_type": "markdown", "source": [ "`Timedelta` — это разница между двумя моментами времени. Отлично подходит для расчета длительности сессий." ], "metadata": { "id": "VxkphzXvejhV" } }, { "cell_type": "code", "source": [ "# Добавим время выхода из игры\n", "df['logout_time'] = pd.to_datetime(['2026-11-08 16:00:00', '2026-11-08 17:15:00', '2026-11-09 11:30:00'])\n", "\n", "# Считаем время в игре\n", "df['session_duration'] = df['logout_time'] - df['login_time']\n", "print(df['session_duration'])\n", "\n", "# Можем перевести в минуты или часы\n", "df['duration_minutes'] = df['session_duration'].dt.total_seconds() / 60\n", "df" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 213 }, "id": "Y9kai7zxelud", "outputId": "c6d3e614-7cee-41a1-934a-6e7de022164d" }, "execution_count": 7, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "0 0 days 01:30:00\n", "1 0 days 01:30:00\n", "2 0 days 01:30:00\n", "Name: session_duration, dtype: timedelta64[ns]\n" ] }, { "output_type": "execute_result", "data": { "text/plain": [ " login_time year month day hour dayofweek day_name \\\n", "0 2026-11-08 14:30:00 2026 11 8 14 6 Sunday \n", "1 2026-11-08 15:45:00 2026 11 8 15 6 Sunday \n", "2 2026-11-09 10:00:00 2026 11 9 10 0 Monday \n", "\n", " logout_time session_duration duration_minutes \n", "0 2026-11-08 16:00:00 0 days 01:30:00 90.0 \n", "1 2026-11-08 17:15:00 0 days 01:30:00 90.0 \n", "2 2026-11-09 11:30:00 0 days 01:30:00 90.0 " ], "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", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
login_timeyearmonthdayhourdayofweekday_namelogout_timesession_durationduration_minutes
02026-11-08 14:30:002026118146Sunday2026-11-08 16:00:000 days 01:30:0090.0
12026-11-08 15:45:002026118156Sunday2026-11-08 17:15:000 days 01:30:0090.0
22026-11-09 10:00:002026119100Monday2026-11-09 11:30:000 days 01:30:0090.0
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "df", "summary": "{\n \"name\": \"df\",\n \"rows\": 3,\n \"fields\": [\n {\n \"column\": \"login_time\",\n \"properties\": {\n \"dtype\": \"date\",\n \"min\": \"2026-11-08 14:30:00\",\n \"max\": \"2026-11-09 10:00:00\",\n \"num_unique_values\": 3,\n \"samples\": [\n \"2026-11-08 14:30:00\",\n \"2026-11-08 15:45:00\",\n \"2026-11-09 10:00:00\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"year\",\n \"properties\": {\n \"dtype\": \"int32\",\n \"num_unique_values\": 1,\n \"samples\": [\n 2026\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"month\",\n \"properties\": {\n \"dtype\": \"int32\",\n \"num_unique_values\": 1,\n \"samples\": [\n 11\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"day\",\n \"properties\": {\n \"dtype\": \"int32\",\n \"num_unique_values\": 2,\n \"samples\": [\n 9\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"hour\",\n \"properties\": {\n \"dtype\": \"int32\",\n \"num_unique_values\": 3,\n \"samples\": [\n 14\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"dayofweek\",\n \"properties\": {\n \"dtype\": \"int32\",\n \"num_unique_values\": 2,\n \"samples\": [\n 0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"day_name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 2,\n \"samples\": [\n \"Monday\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"logout_time\",\n \"properties\": {\n \"dtype\": \"date\",\n \"min\": \"2026-11-08 16:00:00\",\n \"max\": \"2026-11-09 11:30:00\",\n \"num_unique_values\": 3,\n \"samples\": [\n \"2026-11-08 16:00:00\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"session_duration\",\n \"properties\": {\n \"dtype\": \"timedelta64[ns]\",\n \"num_unique_values\": 1,\n \"samples\": [\n \"0 days 01:30:00\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"duration_minutes\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.0,\n \"min\": 90.0,\n \"max\": 90.0,\n \"num_unique_values\": 1,\n \"samples\": [\n 90.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}" } }, "metadata": {}, "execution_count": 7 } ] }, { "cell_type": "markdown", "source": [ "# 🚀 Практика: Аналитика сайта \"Школьные конспекты\"" ], "metadata": { "id": "iM0Lrh8QeyUP" } }, { "cell_type": "markdown", "source": [ "Ты запустил свой сайт для обмена конспектами. У тебя есть сырые данные о том, когда пользователи скачивали материалы.\n", "Твоя задача: очистить данные, извлечь признаки и ответить на бизнес-вопросы." ], "metadata": { "id": "EofFv9kke2Z8" } }, { "cell_type": "code", "source": [ "import pandas as pd\n", "\n", "# Сырые данные (обрати внимание, что даты в разных форматах и есть пропуски/ошибки)\n", "raw_data = {\n", " 'user_id': [1, 2, 3, 4, 5, 6, 7],\n", " 'download_time': [\n", " '2026-11-02 09:15:00', # Понедельник\n", " '03.11.2026 14:30', # Вторник (другой формат)\n", " '2026-11-04 18:00:00', # Среда\n", " 'invalid_date', # Ошибка в данных\n", " '2026-11-07 10:00:00', # Суббота\n", " '2026-11-08 23:45:00', # Воскресенье (сегодня!)\n", " '2026-11-02 11:00:00' # Понедельник\n", " ],\n", " 'subject': ['Math', 'English', 'Physics', 'Math', 'IT', 'IT', 'English']\n", "}\n", "\n", "df = pd.DataFrame(raw_data)\n", "df" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 269 }, "id": "oPsRXrE-e6bc", "outputId": "7b3cefcb-7593-4377-fb7a-205cf25a48a4" }, "execution_count": 8, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ " user_id download_time subject\n", "0 1 2026-11-02 09:15:00 Math\n", "1 2 03.11.2026 14:30 English\n", "2 3 2026-11-04 18:00:00 Physics\n", "3 4 invalid_date Math\n", "4 5 2026-11-07 10:00:00 IT\n", "5 6 2026-11-08 23:45:00 IT\n", "6 7 2026-11-02 11:00:00 English" ], "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", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
user_iddownload_timesubject
012026-11-02 09:15:00Math
1203.11.2026 14:30English
232026-11-04 18:00:00Physics
34invalid_dateMath
452026-11-07 10:00:00IT
562026-11-08 23:45:00IT
672026-11-02 11:00:00English
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "df", "summary": "{\n \"name\": \"df\",\n \"rows\": 7,\n \"fields\": [\n {\n \"column\": \"user_id\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 2,\n \"min\": 1,\n \"max\": 7,\n \"num_unique_values\": 7,\n \"samples\": [\n 1,\n 2,\n 6\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"download_time\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 7,\n \"samples\": [\n \"2026-11-02 09:15:00\",\n \"03.11.2026 14:30\",\n \"2026-11-08 23:45:00\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"subject\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n \"English\",\n \"IT\",\n \"Math\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}" } }, "metadata": {}, "execution_count": 8 } ] }, { "cell_type": "markdown", "source": [ "## Очистка и парсинг дат" ], "metadata": { "id": "9wLAViVhe_nU" } }, { "cell_type": "markdown", "source": [ "Преобразуй столбец `download_time` в формат `datetime`.\n", "Подсказка: Используй `pd.to_datetime(..., errors='coerce')`, чтобы некорректные даты (как 'invalid_date') превратились в `NaT` (Not a Time), а не вызывали ошибку. Удали строки с `NaT`." ], "metadata": { "id": "rkGW6RsbfC3A" } }, { "cell_type": "code", "source": [ "# ТВОЙ КОД ЗДЕСЬ" ], "metadata": { "id": "JdxoxkeHfNz8" }, "execution_count": 9, "outputs": [] }, { "cell_type": "markdown", "source": [ "## Извлечение признаков" ], "metadata": { "id": "hAvklS3gfPqc" } }, { "cell_type": "markdown", "source": [ "Создай новые столбцы:\n", "1. day_of_week (число от 0 до 6)\n", "2. is_weekend (булево значение: True если суббота или воскресенье)\n", "3. hour (час скачивания)" ], "metadata": { "id": "awZMk5eAfXEr" } }, { "cell_type": "code", "source": [ "# ТВОЙ КОД ЗДЕСЬ" ], "metadata": { "id": "SOQLXm__fb4z" }, "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "source": [ "## Бизнес-аналитика" ], "metadata": { "id": "O2K6OK5TfhSc" } }, { "cell_type": "markdown", "source": [ "Ответь на вопросы, используя группировку (groupby) и фильтрацию:\n", "1. В какие дни недели (названия дней) пользователи скачивают конспекты чаще всего?\n", "2. Какой процент скачиваний приходится на выходные (is_weekend)?\n", "3. Во сколько часов (от 0 до 23) пик активности пользователей?" ], "metadata": { "id": "gQ3B95eifl-r" } }, { "cell_type": "code", "source": [ "# ТВОЙ КОД ЗДЕСЬ" ], "metadata": { "id": "bfPZtpRRfp5M" }, "execution_count": null, "outputs": [] } ] }