Mastodon 🐘

Tatu LeppämäkiThank you to <a href="https://mstdn.social/tags/Kone" class="mention hashtag" rel="tag">#Kone</a> & Mai and Tor Nessling Foundations for supporting this work. A quantitative work like this would not be possible without a robust suite of FOSS tools. My thanks to the maintainers of <a href="https://mstdn.social/tags/QGIS" class="mention hashtag" rel="tag">#QGIS</a>, <a href="https://mstdn.social/tags/pandas" class="mention hashtag" rel="tag">#pandas</a>, <a href="https://mstdn.social/tags/geopandas" class="mention hashtag" rel="tag">#geopandas</a>, <a href="https://mstdn.social/tags/duckdb" class="mention hashtag" rel="tag">#duckdb</a>, <a href="https://mstdn.social/tags/dask" class="mention hashtag" rel="tag">#dask</a>, <a href="https://mstdn.social/tags/statsmodels" class="mention hashtag" rel="tag">#statsmodels</a>, <a href="https://mstdn.social/tags/jupyter" class="mention hashtag" rel="tag">#jupyter</a> and many more!

HabrКак пакет с пакетами помог аналитику решить задачу для бизнеса, или keep calm and import statsmodelsВсем привет! Меня зовут Сабина, я лидер команды исследователей данных во ВкусВилле. Мы помогаем бизнесу принимать решения, ориентируясь в том числе на данные. Сегодня я расскажу об одном таком случае. Статья будет полезна аналитикам, которые хотят перестать беспокоиться и начать использовать линейную регрессию из питоновской библиотеки stasmodels.<a href="https://habr.com/ru/companies/vkusvill/articles/851264/" rel="nofollow noopener" translate="no" target="_blank">https://habr.com/ru/companies/vkusvill/articles/851264/</a><a href="https://zhub.link/tags/data_science" class="mention hashtag" rel="nofollow noopener" target="_blank">#data_science</a> <a href="https://zhub.link/tags/python" class="mention hashtag" rel="nofollow noopener" target="_blank">#python</a> <a href="https://zhub.link/tags/statsmodels" class="mention hashtag" rel="nofollow noopener" target="_blank">#statsmodels</a> <a href="https://zhub.link/tags/linear_regression" class="mention hashtag" rel="nofollow noopener" target="_blank">#linear_regression</a> <a href="https://zhub.link/tags/%D0%BB%D0%B8%D0%BD%D0%B5%D0%B9%D0%BD%D0%B0%D1%8F_%D1%80%D0%B5%D0%B3%D1%80%D0%B5%D1%81%D1%81%D0%B8%D1%8F" class="mention hashtag" rel="nofollow noopener" target="_blank">#линейная_регрессия</a>

HabrКак обнаружить и устранить мультиколлинеарность с помощью Statsmodels в ПитонеПривет, Хабр! Мультиколлинеарность возникает, когда в модели множественной регрессии одна из независимых переменных может быть линейно предсказана с помощью других независимых переменных с высокой степенью точности. Это явление приводит к тому, что расчетные коэффициенты регрессии становятся нестабильными и их значения могут сильно изменяться в зависимости от включения или исключения других переменных в модель. Высокая мультиколлинеарность может привести к значительному изменению коэффициентов при незначительных изменениях в данных или спецификации модели. Это усложняет интерпретацию коэффициентов, поскольку они могут значительно изменяться от одного анализа к другому. Когда переменные сильно коррелированы, стандартные ошибки оценок коэффициентов увеличиваются. Это ведет к увеличению p -значений, что может ошибочно привести к заключению о том, что переменные не имеют значимого влияния на зависимую переменную, хотя на самом деле это не так. В статье рассмотрим как обнаружить и устранить мультиколлинеарность с помощью Statsmodels в Питоне.<a href="https://habr.com/ru/companies/otus/articles/810453/" rel="nofollow noopener" translate="no" target="_blank">https://habr.com/ru/companies/otus/articles/810453/</a><a href="https://zhub.link/tags/data_science" class="mention hashtag" rel="nofollow noopener" target="_blank">#data_science</a> <a href="https://zhub.link/tags/ML" class="mention hashtag" rel="nofollow noopener" target="_blank">#ML</a> <a href="https://zhub.link/tags/python" class="mention hashtag" rel="nofollow noopener" target="_blank">#python</a> <a href="https://zhub.link/tags/statsmodels" class="mention hashtag" rel="nofollow noopener" target="_blank">#statsmodels</a>

HabrИндуктивная статистика: доверительные интервалы, предельные ошибки, размер выборки и проверка гипотезОдной из самых распространённых задач современной аналитики является формирование суждений о большой совокупности (например, о миллионах пользователей приложения), опираясь на данные лишь о небольшой части этой совокупности - выборке. Можно ли сделать вывод о миллионной аудитории крупного мобильного приложения, собрав данные об использовании лишь для 100 пользователей? Или стоит собрать данные для 1000 пользователей? Ответ интуитивно прост и понятен: чем больше данных есть в наличии, тем более точными будут прогнозируемые результаты для всей совокупности. Какую вероятность ошибиться при анализе мы можем допустить: 5% или 1%? Относятся ли две выборки к одной совокупности, или между ними есть ощутимая значимая разница и они относятся к разным совокупностям? Точность прогноза и вероятность ошибки при ответе на эти и другие вопросы поддаются вполне конкретным расчётам и могут корректироваться в зависимости от потребностей продукта и бизнеса на этапе планирования и подготовки эксперимента. Рассмотрим подробнее, как параметры эксперимента и статистические критерии оказывают влияние на результаты анализа и выводы обо всей совокупности, а для этого смоделируем тысячу A/A , A/B и A/B/C/D тестов .<a href="https://habr.com/ru/articles/807051/" rel="nofollow noopener" translate="no" target="_blank">https://habr.com/ru/articles/807051/</a><a href="https://zhub.link/tags/%D0%BC%D0%B0%D1%82%D0%B5%D0%BC%D0%B0%D1%82%D0%B8%D0%BA%D0%B0" class="mention hashtag" rel="nofollow noopener" target="_blank">#математика</a> <a href="https://zhub.link/tags/%D0%BC%D0%B0%D1%82%D0%B5%D0%BC%D0%B0%D1%82%D0%B8%D1%87%D0%B5%D1%81%D0%BA%D0%B0%D1%8F_%D1%81%D1%82%D0%B0%D1%82%D0%B8%D1%81%D1%82%D0%B8%D0%BA%D0%B0" class="mention hashtag" rel="nofollow noopener" target="_blank">#математическая_статистика</a> <a href="https://zhub.link/tags/%D0%B0%D0%BD%D0%B0%D0%BB%D0%B8%D0%B7_%D0%B4%D0%B0%D0%BD%D0%BD%D1%8B%D1%85" class="mention hashtag" rel="nofollow noopener" target="_blank">#анализ_данных</a> <a href="https://zhub.link/tags/%D1%81%D1%82%D0%B0%D1%82%D0%B8%D1%81%D1%82%D0%B8%D1%87%D0%B5%D1%81%D0%BA%D0%B8%D0%B9_%D0%B0%D0%BD%D0%B0%D0%BB%D0%B8%D0%B7" class="mention hashtag" rel="nofollow noopener" target="_blank">#статистический_анализ</a> <a href="https://zhub.link/tags/ab_%D1%82%D0%B5%D1%81%D1%82%D1%8B" class="mention hashtag" rel="nofollow noopener" target="_blank">#ab_тесты</a> <a href="https://zhub.link/tags/statsmodels" class="mention hashtag" rel="nofollow noopener" target="_blank">#statsmodels</a> <a href="https://zhub.link/tags/scipy" class="mention hashtag" rel="nofollow noopener" target="_blank">#scipy</a> <a href="https://zhub.link/tags/python" class="mention hashtag" rel="nofollow noopener" target="_blank">#python</a> <a href="https://zhub.link/tags/matplotlib" class="mention hashtag" rel="nofollow noopener" target="_blank">#matplotlib</a> <a href="https://zhub.link/tags/%D0%BF%D1%80%D0%BE%D0%B2%D0%B5%D1%80%D0%BA%D0%B0_%D0%B3%D0%B8%D0%BF%D0%BE%D1%82%D0%B5%D0%B7" class="mention hashtag" rel="nofollow noopener" target="_blank">#проверка_гипотез</a>

pi<a href="https://mastodon.sdf.org/tags/Python" class="mention hashtag" rel="nofollow noopener" target="_blank">#Python</a> in <a href="https://mastodon.sdf.org/tags/Excel" class="mention hashtag" rel="nofollow noopener" target="_blank">#Excel</a> (in Beta) <a href="https://mastodon.sdf.org/tags/Microsoft" class="mention hashtag" rel="nofollow noopener" target="_blank">#Microsoft</a> 🤝 <a href="https://mastodon.sdf.org/tags/Anaconda" class="mention hashtag" rel="nofollow noopener" target="_blank">#Anaconda</a> Default imported libraries: <a href="https://mastodon.sdf.org/tags/matplotlib" class="mention hashtag" rel="nofollow noopener" target="_blank">#matplotlib</a> <a href="https://mastodon.sdf.org/tags/numpy" class="mention hashtag" rel="nofollow noopener" target="_blank">#numpy</a> <a href="https://mastodon.sdf.org/tags/pandas" class="mention hashtag" rel="nofollow noopener" target="_blank">#pandas</a> <a href="https://mastodon.sdf.org/tags/seaborn" class="mention hashtag" rel="nofollow noopener" target="_blank">#seaborn</a> <a href="https://mastodon.sdf.org/tags/statsmodels" class="mention hashtag" rel="nofollow noopener" target="_blank">#statsmodels</a>only for Windows, needs internet access, code executed on MS servers without network or file accesssee <a href="https://aka.ms/python-in-excel-getting-started" rel="nofollow noopener" target="_blank">https://aka.ms/python-in-excel-getting-started</a> & <a href="https://www.anaconda.com/excel" rel="nofollow noopener" target="_blank">https://www.anaconda.com/excel</a>

Tim Kelloggoh wow! You'll be able to use <a href="https://hachyderm.io/tags/python" class="mention hashtag" rel="nofollow noopener" target="_blank">#python</a> from within <a href="https://hachyderm.io/tags/excel" class="mention hashtag" rel="nofollow noopener" target="_blank">#excel</a> and <a href="https://hachyderm.io/tags/powerquery" class="mention hashtag" rel="nofollow noopener" target="_blank">#powerquery</a> soon. And that Python install includes <a href="https://hachyderm.io/tags/pandas" class="mention hashtag" rel="nofollow noopener" target="_blank">#pandas</a>, <a href="https://hachyderm.io/tags/matplotlib" class="mention hashtag" rel="nofollow noopener" target="_blank">#matplotlib</a> and <a href="https://hachyderm.io/tags/statsmodels" class="mention hashtag" rel="nofollow noopener" target="_blank">#statsmodels</a> <a href="https://www.theverge.com/2023/8/22/23841167/microsoft-excel-python-integration-support" rel="nofollow noopener" target="_blank">https://www.theverge.com/2023/8/22/23841167/microsoft-excel-python-integration-support</a>

Juan LuisNoticias sobre Python y Datos de la semana, episodio 76 🐍⚙️En resumen: Edición ultrarrápida de domingo por la tarde: versiones nuevas de numba, statsmodels y geopandas, por qué no usar leyendas en matplotlib, y primeras pruebas con StarCoder.<a href="https://buttondown.email/astrojuanlu/archive/episodio-76/" rel="nofollow noopener" target="_blank">https://buttondown.email/astrojuanlu/archive/episodio-76/</a>Apoya el noticiero suscribiéndote por correo 📬<a href="https://social.juanlu.space/tags/noticieropythonydatos" class="mention hashtag" rel="nofollow noopener" target="_blank">#noticieropythonydatos</a> <a href="https://social.juanlu.space/tags/python" class="mention hashtag" rel="nofollow noopener" target="_blank">#python</a> <a href="https://social.juanlu.space/tags/pydata" class="mention hashtag" rel="nofollow noopener" target="_blank">#pydata</a> <a href="https://social.juanlu.space/tags/numba" class="mention hashtag" rel="nofollow noopener" target="_blank">#numba</a> <a href="https://social.juanlu.space/tags/statsmodels" class="mention hashtag" rel="nofollow noopener" target="_blank">#statsmodels</a> <a href="https://social.juanlu.space/tags/geopandas" class="mention hashtag" rel="nofollow noopener" target="_blank">#geopandas</a> <a href="https://social.juanlu.space/tags/matplotlib" class="mention hashtag" rel="nofollow noopener" target="_blank">#matplotlib</a> <a href="https://social.juanlu.space/tags/starcoder" class="mention hashtag" rel="nofollow noopener" target="_blank">#starcoder</a> <a href="https://social.juanlu.space/tags/polars" class="mention hashtag" rel="nofollow noopener" target="_blank">#polars</a> ¡Sigue a <a href="https://fosstodon.org/@numba" class="u-url mention" rel="nofollow noopener" target="_blank">@numba</a> y <a href="https://fosstodon.org/@geopandas" class="u-url mention" rel="nofollow noopener" target="_blank">@geopandas</a> en Mastodon!

ChristopherI just did an <a href="https://tenforward.social/tags/introduction" class="mention hashtag" rel="nofollow noopener" target="_blank">#introduction</a> a few days ago, but I've moved servers, so let's try one more time, for the cheap seats in the back!I'm currently a data analyst/product <a href="https://tenforward.social/tags/DataScientist" class="mention hashtag" rel="nofollow noopener" target="_blank">#DataScientist</a> working with free-to-play <a href="https://tenforward.social/tags/VideoGames" class="mention hashtag" rel="nofollow noopener" target="_blank">#VideoGames</a>, and living in <a href="https://tenforward.social/tags/Halifax" class="mention hashtag" rel="nofollow noopener" target="_blank">#Halifax</a>, <a href="https://tenforward.social/tags/NovaScotia" class="mention hashtag" rel="nofollow noopener" target="_blank">#NovaScotia</a>, <a href="https://tenforward.social/tags/Canada" class="mention hashtag" rel="nofollow noopener" target="_blank">#Canada</a>. I've done a lot of work on <a href="https://tenforward.social/tags/Analytics" class="mention hashtag" rel="nofollow noopener" target="_blank">#Analytics</a> design, with a focus on ensuring player telemetry events are sensibly cross-referenceable, and looking for relationships between engagement with different game features and business outcomes. Business teams in freemium games love looking for magic buttons.I primarily use <a href="https://tenforward.social/tags/SQL" class="mention hashtag" rel="nofollow noopener" target="_blank">#SQL</a>, <a href="https://tenforward.social/tags/Pandas" class="mention hashtag" rel="nofollow noopener" target="_blank">#Pandas</a>, <a href="https://tenforward.social/tags/Statsmodels" class="mention hashtag" rel="nofollow noopener" target="_blank">#Statsmodels</a>, and <a href="https://tenforward.social/tags/SKLearn" class="mention hashtag" rel="nofollow noopener" target="_blank">#SKLearn</a> on <a href="https://tenforward.social/tags/Databricks" class="mention hashtag" rel="nofollow noopener" target="_blank">#Databricks</a> (<a href="https://tenforward.social/tags/Python" class="mention hashtag" rel="nofollow noopener" target="_blank">#Python</a>), and <a href="https://tenforward.social/tags/JuliaLang" class="mention hashtag" rel="nofollow noopener" target="_blank">#JuliaLang</a> (DataFrames.jl, GLM.jl, Gadfly.jl, and Makie.jl, etc) for smaller, locally run projects. My interests lie in expanding the library of ML models I have in my back pocket for performing inference based knowledge generation. I'm not super keen on automating products with quasi-black-boxes for the sake of revenue optimization. If I'm not personally learning something new about people through my work, I don't usually see the value in it.I did my BSc in <a href="https://tenforward.social/tags/Physics" class="mention hashtag" rel="nofollow noopener" target="_blank">#Physics</a> and my MSc in <a href="https://tenforward.social/tags/Astronomy" class="mention hashtag" rel="nofollow noopener" target="_blank">#Astronomy</a>, and, though I had dreams of progressing further down that pipeline, life kind of got in the way. Between the two degrees, I worked at the <a href="https://tenforward.social/tags/Edmonton" class="mention hashtag" rel="nofollow noopener" target="_blank">#Edmonton</a> <a href="https://tenforward.social/tags/Planetarium" class="mention hashtag" rel="nofollow noopener" target="_blank">#Planetarium</a> for four years as a presenter/operator (should out to the <a href="https://tenforward.social/tags/ZeidlerDome" class="mention hashtag" rel="nofollow noopener" target="_blank">#ZeidlerDome</a> at <a href="https://tenforward.social/tags/TWoSE" class="mention hashtag" rel="nofollow noopener" target="_blank">#TWoSE</a>!). I'm a life-long <a href="https://tenforward.social/tags/Trekie" class="mention hashtag" rel="nofollow noopener" target="_blank">#Trekie</a>, thanks to my mother. I grew up with <a href="https://tenforward.social/tags/TNG" class="mention hashtag" rel="nofollow noopener" target="_blank">#TNG</a> and <a href="https://tenforward.social/tags/DS9" class="mention hashtag" rel="nofollow noopener" target="_blank">#DS9</a>, and watched the first 5 seasons of <a href="https://tenforward.social/tags/VOY" class="mention hashtag" rel="nofollow noopener" target="_blank">#VOY</a> before leaving home for university. Currently very bullish on <a href="https://tenforward.social/tags/SNW" class="mention hashtag" rel="nofollow noopener" target="_blank">#SNW</a> and <a href="https://tenforward.social/tags/LDS" class="mention hashtag" rel="nofollow noopener" target="_blank">#LDS</a>. I'm also a lifelong <a href="https://tenforward.social/tags/Baseball" class="mention hashtag" rel="nofollow noopener" target="_blank">#Baseball</a> fan (<a href="https://tenforward.social/tags/BlueJays" class="mention hashtag" rel="nofollow noopener" target="_blank">#BlueJays</a> and <a href="https://tenforward.social/tags/Expos" class="mention hashtag" rel="nofollow noopener" target="_blank">#Expos</a>), and actively play rec <a href="https://tenforward.social/tags/Softball" class="mention hashtag" rel="nofollow noopener" target="_blank">#Softball</a>. About a year ago, I purchased my first lens-swappable digital <a href="https://tenforward.social/tags/camera" class="mention hashtag" rel="nofollow noopener" target="_blank">#camera</a>, and have been figuring out <a href="https://tenforward.social/tags/Photography" class="mention hashtag" rel="nofollow noopener" target="_blank">#Photography</a> ever since. Most of my posts have focused on sharing my pictures, though I've recently decided to start a dedicated account for that on a <a href="https://tenforward.social/tags/PixelFed" class="mention hashtag" rel="nofollow noopener" target="_blank">#PixelFed</a> server, for the sake of searchability.My wife is currently studying political sociology, and I find her work fascinating. She's not currently on the Fediverse, but maybe one of these days. This has really lost all sense of narrative flow, hasn't it? Oops!

Recent searches

Search options

Administered by:

Server stats:

#Statsmodels