Amazon AWS Data Pipeline

Ok I missed this one as it came on Dec 20.  I think the AWS data pipeline is a really important step forward for cloud enabled analytics.

dp-how-dp-works-v2

http://aws.amazon.com/about-aws/whats-new/2012/12/20/announcing-aws-data-pipeline/

What is AWS Data Pipeline?

AWS Data Pipeline is a web service that you can use to automate the movement and transformation of data. Continue reading “Amazon AWS Data Pipeline”

Funny Stuff on Book Resellers and Dynamic Pricing at Amazon

I have been keeping an eye on the pricing figure of R for Business Analytics at Amazon.com and have watched it fluctuate between 49$ to 59$ while it is currently at 50$. (Incidentally- Per Capita GDP of India is 3700$ per year or ~$10 per day)

My point is

  1. why should a used book reseller offer to sell the same book at a HIGHER price,
  2. while a new book reseller offers to sell the same book that Amazon is selling (in stock) at a higher price.
  3. On top of that, these resellers have hundreds of thousands of ratings from delighted book buyers (of other books) who were very happy at buying the same book (used or new) at higher prices including upto twice the price.

Funny Stuff , huh!

Continue reading “Funny Stuff on Book Resellers and Dynamic Pricing at Amazon”

Writing a technical book

This is a fairly concise collection on how to write a technical book. It may seem arrogant for a 1- book author like me to do so, but I get a lot of queries on this and it seems there is a fair amount of information asymmetry on this process.  I have experience with getting rejected and accepted in both creative and technology domains, but I will make this post fairly tech specific.

Books I have Written-(click on images to go to the book site)

Cred-

Poetry (Self Published)

In Case I Don't See You Again
Corporate Poetry
Poets & Hackers (e-book)
Technology (Published )
R for Business Analytics
(Currently Writing)
R for Cloud Computing ( Springer) – Due 2013
R for Web Analytics and Social Media Analytics (Springer) – Due 2014
Top 5 Myths on Writing and Getting Published
  • Publishers dont like unsolicited manuscripts.

Well they don’t like unsolicited manuscripts from total unknowns. This is also very domain specific. If you are writing a novel, or a poetry book, or a technical book, approval rates will depend on current interest in that domain.

Advice– If you are first time author to be, choose your niche domain as one which you are passionate about and which has been generating some buzz lately. It could be Python, D3, R etc.

  • Publishers get all the money

No, they don’t make that much money compared to a Hollywood studio. Yes, books are expensive, but they basically are funding a whole supply chain that may or may not be efficient. Your book is subsidizing all the books that didn’t sell. Proof reading, and editing are not very glamorous jobs, but they take a long time, and are expensive. I have much more respect for editors now than say 3 years ago. The ultimate in supply chain efficiency would be if each and every hard copy was printed on demand, and each and every soft copy was priced efficiently given pricing elasticity. Pricing analytics on dynamic book pricing (like on Amazon)— hmm

  • Writers get all the money

You would be lucky to get more than 14% from a gross selling price of a hard copy or more than 40% of an electronic book. You want to make money, dont write technical books, write white papers and make webinars.

  • Writers get no money

You don’t make money by writing a technical book, but your branding does go up significantly, and you can now charge for training, webinars, talks, conferences, white papers, articles. These alternatives can help you survive.

  • I got a great idea- but I keep getting rejected. That guy had a lousy idea, but he keeps writing.

THAT guy wrote a great proposal, spent time building his brand, and wrote interesting stuff. Publishers like to sell books, not ideas.Writer jealousy and insecurity are part of the game – you have a limited amount of energy in a day- spend that writing or spend that reading. Ideally do both.

Book Publication

The book publication process has three parts-

1) Proposal

2) Manuscript

3) Editing

1) Proposal- Write an awesome proposal. Take tips from the publisher website. Choose which publisher is more interested in publishing the topic (hint- go to all the websites) . Those publisher websites confusing you yet- jump to the FAQ.

Some publishers I think relevant to technical books-

http://www.springer.com/authors/book+authors/faq+for+book+authors?SGWID=0-1725014-0-0-0

http://support.sas.com/community/authors/index.html

http://oreilly.com/oreilly/author/intro.csp

2) Manuscript- Write daily . 300 words. 300 times. Thats a manuscript. It is tough for people like us. Hemingway had  it easy. I used a Latex GUI called Lyx for writing http://www.lyx.org/. You may choose your own tool, style, time of day /night, cafe , room to spur your creative juices.

3) Editing- you will edit, chop, re edit and rewrite a book many times. It is ok. Make it readable is my advice. Try and think of a non technical person and try and explain your book to clear your ideas.

Once your proposal is accepted, you sign a contract for royalty and copyright.

Once the contract is signed you write the manuscript.This also involves a fair amount of research, citations, folder management , to keep your book figures, your citations ready. I generally write the citation then and there within the book, and then organize them later chapter by chapter. Un-cited work leads to charges of plagiarism which is the buzz kill for any author. Write, Cite, Rewrite.

You will also need to create index (can be done by software) so people can navigate the book better , and appendix for hiding all the stuff you couldn’t leave behind.

Once you submit the manuscript ,you choose the cover, discuss the rewrites with editor, edit the changes suggested, and resend the manuscript files, count till six months for publication. Send copies to people you like who can help spread the word on your book. Wait for reviews, engage with positivity with everyone, then wait for sales figures. Congrats- you are a writer now!

 

 

 

Amazon drops prices of Linux AMIs by ~20%

Amazon cloud gets more exciting. We are still waiting for the Oracle and Google public clouds (compute) to open up out of beta! See their (rather cluttered) blog

http://aws.typepad.com/aws/2012/10/new-ec2-second-generation-standard-instances-and-price-reductions-1.html

Today, we are excited to announce a new generation of the original Amazon EC2 instance family. Second generation Standard instances (M3 instances) provide customers with the same balanced set of CPU and memory resources as first generation Standard instances (M1 instances) while providing customers with 50% more computational capability/core.

M3 instances are currently available in two instance types; extra-large (m3.xlarge) and double extra-large (m3.2xlarge). Examples of applications that can benefit from the additional CPU horsepower of these new instances include media encoding, batch processing, web servers, caching fleets, and many others. Currently, M3 instances are available in the US East (N. Virginia) Region starting at a Linux On-Demand price of $0.58/hr for extra-large instances. Customers can also purchase M3 instances as Reserved Instances or as Spot instances. We will introduce M3 instances in additional regions in the coming months.

To learn more about Amazon EC2 instance types and to find out which instance type might be useful for you, please visit the Amazon EC2 Instance type page.

Pricing Change for M1 Standard Instances
Along with the introduction of the M3 Standard instance family, we are announcing a reduction in Linux On-Demand pricing for M1 Standard instances in the US East (N. Virginia) and US West (Oregon) Regions by almost 19%. The new pricing is effective from November 1 and is described in the following table

Instance Type Previous Price New Price
m1.small $0.080 $0.065
m1.medium $0.160 $0.130
m1.large $0.320 $0.260
m1.xlarge $0.640 $0.520

You can find out more about pricing for all Amazon EC2 instances by visiting the Amazon EC2 pricing page.

 

Databases in the cloud

One more day of me mucking around MySQL and Amazon (hoping to get to the R)

R now part of Amazon Linux AMI

Based on this post, Amazon now had decided to bundle R with Amazon Linux AMI

http://aws.typepad.com/aws/2012/10/amazon-linux-ami-201209-now-available.html

R 2.15: Also coming from your requests, we have added the R language to the Amazon Linux AMI.  We are here to serve your statistical analysis needs!  Simply yum install R and off you go.

ps- back to work. sorry for the delayed posts . I am working on book 2 for Springer- “R for Cloud Computing” . If you have any case studies of R on Amazon,Google, Oracle or Azure clouds please let me know.

pps- With 48 mb, is R too big to bundle in the many default Linux distros . Thoughts?