Curso para Data science

Edx me mando un mail sobre cursos para Data science y yo como soy tan bueno, decidí compartirlo:

Courses | My Account It's not news that data science is one of the hottest fields today. Now is the time to jump in! Whether you're just starting out or looking to up your game, check out the courses below and start data wrangling today. Is this you? You know data scientists are in-demand and you're interested in pursuing a career — but don't have the needed skills or professional background. With minimal to no prerequisite knowledge required, these courses are a great place to start. You'll quickly build foundational data science skills to jumpstart your career. HarvardX Data Science: R Basics Enroll Now HarvardX Data Science: Visualization Enroll Now ColumbiaX Statistical Thinking for Data Science and Analytics Enroll Now ColumbiaX Machine Learning for Data Science and Analytics Enroll Now Microsoft Introduction to Data Science Enroll Now Microsoft Introduction to Python for Data Science Enroll Now Is this you? You have some knowledge and skills in data or computer science and maybe even work in the field. Now you’re looking to up your game to advance your career or pursue an advanced degree. Start your journey from apprentice to expert with these Master's-level courses. UCSanDiegoX Python for Data Science Enroll Now USMx UMUC Cloud Computing for Enterprises Enroll Now AdelaideX Programming for Data Science Enroll Now AdelaideX Computational Thinking and Big Data Enroll Now View All edX Courses edX for Business - eLearning Solutions for Your Company © 2012-2018 edX Inc. EdX, Open edX, and MicroMasters are registered trademarks of edX Inc. View on Web | Unsubscribe 141 Portland St. 9th Floor, Cambridge, MA 02139

Apache Avro

Apache Avro es un sistema de serialización de datos.

Avro proporciona:

• Estructuras de datos ricas
• Un formato de datos binario compacto, rápido.
• Un archivo contenedor, para almacenar datos persistentes.
• Llamada a procedimiento remoto (RPC).
• Integración simple con lenguajes dinámicos. No se requiere la generación de código para leer o escribir archivos de datos ni para usar o implementar protocolos RPC. 

Avro se basa en esquemas. Cuando se leen datos Avro, el esquema utilizado al escribirlo siempre está presente. Esto permite que cada dato se escriba sin gastos generales por valor, lo que hace que la serialización sea rápida y pequeña. Esto también facilita el uso con lenguajes dinámicos de scripting, ya que los datos, junto con su esquema, son completamente autodescriptivos.
Cuando los datos de Avro se almacenan en un archivo, su esquema se almacena con él, de modo que los archivos pueden ser procesados posteriormente por cualquier programa. Si el programa que lee los datos espera un esquema diferente, esto se puede resolver fácilmente, ya que ambos esquemas están presentes.

Cuando se usa Avro en RPC, el cliente y el servidor intercambian esquemas en el enlace de conexión. (Esto se puede optimizar para que, en la mayoría de las llamadas, no se transmitan realmente esquemas.) Dado que tanto el cliente como el servidor tienen el esquema completo del otro, la correspondencia entre los mismos campos con nombre, campos faltantes, campos adicionales, etc. puede resolverse fácilmente .

Los esquemas Avro se definen con JSON. Esto facilita la implementación en Lenguajes que ya tienen bibliotecas JSON. Avro proporciona una funcionalidad similar a sistemas como Thrift, Protocol Buffers, etc. Avro difiere de estos sistemas en los siguientes aspectos fundamentales.

• Tipado dinámico: Avro no requiere que se genere ese código. Los datos siempre van acompañados de un esquema que permite el procesamiento completo de esos datos sin generación de códigos, tipos de datos estáticos, etc. Esto facilita la construcción de sistemas e lenguajes genéricos de procesamiento de datos.
• Datos no etiquetados: dado que el esquema está presente cuando se leen los datos, es necesario codificar considerablemente menos información de tipo con los datos, lo que da como resultado un tamaño de serialización más pequeño.

No hay identificadores de campo asignados manualmente: cuando un esquema cambia, tanto el esquema antiguo como el nuevo siempre están presentes cuando se procesan los datos, por lo que las diferencias se pueden resolver simbólicamente, usando los nombres de los campos.

Dejo link: https://avro.apache.org/

Vue Mastery

Quiero compartir este mail sobre la pagina Vue Mastery dado que hay muchos recursos del framework Vue.js, y nada más ...

We have brand new content for you. Can't see images? Click here... The other day someone asked me:   "How in the world can you afford to give the Vue Community so much free content?" There is the free Vue Intro course, free VueConf videos, free Vue cheat sheet, and the free official Vue News Podcast.  There's a lot. Well, the day has come where I hope that you might help us (and the Vue.js project itself) with a monthly subscription to make it all sustainable.   What do you get? First and foremost, $5 of your monthly subscription we give directly to the core Vue.js project. Secondly, we have a plan to release one video a week to help on your path to Vue Mastery, and I just released the first lesson on myAdvanced Components course (see below).  Only subscribers get access to this content. Thirdly, you help us continue to be able to afford to do the free things we do for the community (see above). Click Here to Get Started Advanced Components Course In this course it is my goal to help you get better at using the full functionality of Vue, scaling and debugging your applications, extending Vue core functionality, and perhaps even contributing to the Vue project itself.   We start by uncovering the secrets of Vue's reactivity system.   Subscribe & Watch Video 🎥 Newest Features in Vue + Q&A with Evan You Join 400+ people who have signed up for a free guided tour of the newest features of Vue.js and get your questions answered by the man who created it, Evan You. 17 VueConf Lightning Talks Last week we polished and released all 17 lightning talks from VueConf US.  These are all free.  From Adam Jahr talking about 5 Libraries you should know about, to Sean "Webpack" Larkin talking about Code Splitting patterns in Vue, there's something here for everyone. We even had Derick Sozo watch them all and write up an article on5 lightning talks from VueConf.US you can't miss.  Yup, free. 🎧 Oh, and a new Podcast Adam and I cover the latest news in the Vue.js community in just 5 minutes, so you can listen in your car, on the train, or when you're washing your cat.  Yup, it's free too. You can listen to it here. Or subscribe to it below. 🍏 Apple Podcasts, 🤖 Android, Stitcher, or RSS Not going to be paying for a subscription? Totally understand if you're not in a position to help us out right now, we're still going to be producing lots of free stuff.  Three things you could do instead if you wanted to be nice: 1. Ask your boss to pay for a subscription, and get us in touch.  2. Help us promote! Forward this email or retweet this tweet. 3. Respond to this email and let us know what we could do to get you to subscribe. Thanks for reading! Gregg Pollack Founder of Vue Mastery Vue Mastery Website You received this email because you wanted to start your journey to Vue Mastery.  If you aren't interested in learning Vue, feel free to: Unsubscribe

Machine Learning Yearning

Como les conté en un post anterior. Andrew Ng esta sacando un libro sobre machine learning.

Me llego otro mail del amigo Andrew Ng sobre su nuevo libro:

View this email in your browser

AI is the new electricity

Machine Learning Yearning Dear friends, Last week, we learned about specific techniques to address avoidable bias and variance. This week’s chapters focus on learning curves. They are an even more informative and visual way to help you figure out how much error can be attributed to avoidable bias or variance. Read this week’s chapters to learn more! Read Chapters 28-30 In Case You Missed It I recently spoke with Greylock's Sarah Guo about how industries can leverage AI technology and the impacts of AI on the future workforce. Listen to the full interview Pun of the Week If you have a great AI pun, tweet it to me @AndrewYNg using #AIpun. I'll share my favorite in next week's email!

Catch up on Machine Learning Yearning Miss last week's email? Access all chapters below:   Ch. 1 - 14 Ch. 15 - 19 Ch. 20 - 22 Ch. 23 - 27 Copyright © 2018 Andrew Ng, All rights reserved. You are receiving this because you opted in to receiving emails about Andrew's upcoming book. Our mailing address is: Andrew Ng 353 Serra Mall Stanford, CA 94305 Add us to your address book

Dstream

Dstream es la abstracción básica en Spark Streaming y representa una stream continuo de datos.

DStream se puede crear a partir de flujos de datos de entrada procedentes de fuentes como Kafka, Flume y Kinesis, o aplicando operaciones en otros DStream. Internamente, un DStream se representa como una secuencia de objetos RDD.

Similar a RDD el DStream soporta:

• map 
• flatMap 
• filter 
• count 
• reduce 
• countByValue 
• reduceByKey 
• join 
• updateStateByKey