2014 Journal

Created: Feb 20, 2018 by Pradeep Gowda Updated: Oct 31, 2023

See also: log / 2015

January

Wednesday, 15 January 2014

Discussed the subject of my MS thesis with my adviser. The topic is “energy aware indoor tracking”. There is growing research and development in the area of Indoor Positioning Systems (IPS) and Indoor tracking. Many of these initiatives make use of sensors available in mobile devices. However, the raising number of sensors and always-on connectivity plays a havoc on battery life of the devices. Understanding indoor tracking with energy usage in context will be my research direction.

Thursday, 16 January 2014

I’m slowly getting into the habit of practicing GTD. Collecting thoughts, TODOs was always the easy part. The recent progress has come in the way of reviewing the material I collect. Thinking “review” as “cleanup” has definitely helped.

Friday, 17 January 2014

“You can only trick them so long. Sooner or later, you’ve got to punch them in the face” – Phil Simms paraphrasing Mike Ditka [in the context of the running game in American Football]. I read this quote as a reminder that you can only use tricks to achieve results only for so long. Ultimately, you will have to return the fundamentals of the game (of whatever variety you are playing). American football is about controlling the line of scrimmage and running the ball down the hill, no matter how much the modern game rules have been altered to suite the more popular passing game.

You are what you (want to) read

If the kind of books you own make you what you aspire to be, this is what I look like (by looking around my study room):

• Distributed systems
• Algorithms
• Programming Langauges Theory
• Datamining and machine learning
• Programming language manuals - C, C++, Java, Lua, tcl, ML, F#, OCaml, Erlang, Lisp, Scheme, Racket, FORTH
• Mathematics - Algebra, Probability, Calculus, History of
• Statistics
• Practice of Programming
• Fitness - weights
• Textbooks from grad school
• Psychology - learning, success, meditation
• Management - time management
• Finance - investing, mutual funds
• Kannada - fiction and non-fiction
• English - non-fiction

Saturday, 18 January 2014

When you claim a certain familiarity, even competence in a field, you begin to notice how narrow and one-dimensional the general reporting appears to be. One can even argue that even classrooms are not immune to this. Consider that a typical engineering school graduate, depending on their school and major is likely to come out thinking that C/Java/.NET is all of programming; Windows is the OS etc.,. This is a distressing thought. So, when I see a “for the masses” production of a specialised field, say - psychology, finance etc., I always warn myself not to take it at face value. The thought process can be summed up as:

“Remember: if the code they show in movies can be so crappy, how accurate can this be?”.

There are many excellent EWDs, but my favourite is EWD1055A – (advice to a young scientist).

Sunday, 19 January 2014

Started reading Doing Data Science in line with this reading strategy I outlined yesterday. The notes can be found here. Steady and slow progress, but very satisfying to deeply read the text.

Thursday, 23 January 2014

A strategy for the Autodidact

Today, between the MOOCs, free and open licensed text books, class notes online and a steady stream of recommendations via twitter, prismatic etc., it is quite easy to drown in a sea of “i want to study this, but there is too much good stuff …”

My problem is that I have a lot of good textbooks in physical form. And having a book take up space in your study room is a good and painful reminder that you haven’t read that book. Not reading a book that you bought 10 years ago is quite embrassing.

So, one solution is to find courses, that use the text book you have on your hand. Naturally the slides and lecture notes capture only a small portion of the prescribed text. Moreover, the notes capture the highlights and often lack the flow, references etc., that is to be expected in a text book.

Saturday, 25 January 2014

“Genius is simply patience carried to the extreme.” – Georges-Louis Leclerc, Comte de Buffon

Sunday, 26 January 2014

But I think when you still enjoy the preparation, the work part of it, I think you probably still aught to be doing that. I think as soon as I stop enjoying it, and I can’t produce, and I can’t help the team, that’s when I’ll stop playing. If that’s next year, maybe it is, but I certainly want to keep playing. – PFM

Tuesday, 28 January 2014

Problem vs Exercises is an important distinction that one needs to know when studying (esp. mathematics).

Some interesting lines:

• The excercise serves to drill a student in some technique or procedure, and requires little, if any, original thought.
• To be suitable, a problem must be such that the student cannot solve it immediately. One does not complain about a problem being too difficult, but rather too easy.
• And it is worth noting, the more problems one plays with, the more problems one may be able to pose on one’s own. The ability to propose significant problems is one requirement to be a creative mathematician.
• Our definition of an exercise is that you look at it and know immediately how to complete it.
• … by a problem, we mean a more intricate question for which at first one has probably no clue to how to approach it, but by perseverance and inspired effort one can transform it into a sequence of exercises. rb

From repetition and practice

“No head” for mathematics” nearly always means “Will not use a pencil.” – Arthur Latham Baker, Elements of Solid Geometry (1894), page ix.

Wednesday, 29 January 2014

Ramachandra Guha’s article on Shivarama Karanth

I woke up about 5AM to hear Siri reciting 2’s multiplication table in her dream. My Amma always used to say to her friends that I could recite the 60 samvatsaras if she woke me in the middle of the night and asked me recite them. Memorising key things is given importance in the early education and I believe it is beneficial in moderate amounts.

IndyBigData – meetup notes

Books

• Bad Data Handbook pub by Oreilly - 2012.
• Practical Data Science with R pub by Manning [EAP].

Looking for bad data

DBPedia is the data source used for bad data analysis. The data set used - Person.

Discussion

• The definition of bad data has changed.
• Spoofing of User Agent string. “Fractal dimensions”.
• Data analyst should be close to the business.
• Type I and Type II errors
• Sampling will not catch all of bad data

Scavenger Hunt

Can you find:

• example of missing values
• example of clearly
• tbd…

Saturday, 01 February 2014

I should pay more attention to multi-core chips and computing. Joe Armstrong gives cause to be excited for such machines in the future.

It seems like chip designers and language designers live in different parallel universes. Chip designers gleefully add weird instructions to their CPUs that nobody has a clue what to do with, and the leave out the instructions we really want.

That is a such a sad thing.

Tomorrow is Superbowl and it is going to be followed 8 months of no-football, which is quite ..sigh..

February

Sunday, 02 February 2014

A sad horse. Manning’s new team lost the SuperBowl. Sigh.

Monday, 03 February 2014

Extreme use of templating, huh?. I kid. This is an impressive implementation of algorithmic writing for commercial uses.

Tuesday, 04 February 2014

A minute spent here and there is doesn’t feel like much. It’s only when you start thinking about what can be done in units of time, you start corralling the loose minutes into usesable blocks.

Of all the MOOCs i’ve registered, Statlearning from Stanford is the one I’m serious about completing. The course required 3-4 hours of a week. I do have this time. Not in single minutes, but in block of 20/30 minutes a day. There is a danger is assuming that I will be able to find this 3-4 hour block of time over a weekend. However, invariably i will find other more pressing personal engagments.

One way to squeeze these blocks of time out of a regular day is a) always be looking for an opportunity to sneak-in the work (on starlearning in this case). b) attempt to extend that sneak to last a 20-30 minute interval.

Think of this in light of how you casually watches TV. You plop in front of the TV with the remote in hand. You start flipping between channels, something catches your eye for a few minutes, or may be not .. in the end, you realise you have spent over an hour just watching “nothing”.

If you can trick yourself into passivity, which often results in dismay, you might as well trick yourself into the habit of converting 2 minute “headline scanning” sneak into a 30 minutes study session.

The turpentine effect

An interesting read from Venkatesh Rao.

In Overachievement, John Eliot talks about a “training” mindset and a “performance” mindset. The former involves meta-cognition and continuously monitoring your own performance. The latter involves an ability to shut off the meta-cognition and just get lost in “doing”. Great teachers were probably great learners. Great doers may be slower learners, but are great at shutting off the meta-cognition.

It’s clear that I need to amp up the “doing” mode.

No artistic decisions usually means doing everything rather than doing nothing. Artists mainly decide what not to do.

Which is forced upon you when under duress of a deadline. Why not simulate it all the time?

Wednesday, 05 February 2014

Do not end the week with nothing – I can attest to the efficacy of this approach at work. Since I pushed for and started having “demo Fridays”, which features demos of product development on Friday afternoons, I can see the urgency to “wrap up” loose ends to be ready for the demo.

Salient points:

Work you can show off, though, is prima facie evidence of your skills. More prosaically, establish an expectation early that you’re simply going to talk about what you’re doing. If you cannot build things you can show at work, you should build things you can show outside of work. optimize for impact and visibility. Prefer To Work On Things You Can Keep. Consumption Is Sometimes Valuable, But Creation Moves You Forward.

Monday, 10 February 2014

“do a huge volume of work. put yourself on a deadline. it’ll take a while. fight your way through that” – Ira Glass

Tuesday, 13 February 2014

Attended the Indy bigdata conference. Many companies and people trying to figure out what it is. The people who have tried things successfully (eg: NGS) agree with my opinions on the topic – it’s not a solution you can buy pre-packaged; put people and process over technology consideration; do a POC; iterate and improve. A personal goal: be a speaker on the 2015(apr 3) edition.

Wednesday, 12 February 2014

The little girl from the 1981 LEGO ad. One of my all time favourite print ads. I saw this around 2010-11 and as a father of a little girl, the ad really speaks to me. Her name is Rachel Giordano, she is 37 years old, and she’s a practicing naturopathic doctor in Seattle, Washington.

Think of new questions

Take time everyday to think about questions. It is the interesting questions that lead you to think in new ways. Answer questions as a practice. Be in situations where you are forced to answer hard questions. This goes back to my earlier thoughts on using “exams” to improve yourself.

Do not try to go from “I know this, this and this; now what can I do?”. Instead, “This is an interesting/challenging/rewarding/business problem. So, how do I solve this?”

What is the nature of these questions? There must be a “template” for generating new and interesting questions. For instance:

• what is this __ is going?

You should always be able to answer the question: “What did you do today?/ this week?/since our last meeting?”. This is a good way to keep yourself oriented towards a) working towards showable results b) thinking about what you were going to say refines which parts of the task(s) you focus on.

Also, to be able to answer this with certainty and vigour, you must know that one outcome that you are pouring all your energy into.

How do you say?

Many weeks ago, I thought up an idea for a website, which I thought I would call “how do you .com”, which, as the name implies, plays back a audio recording of a word you type into a search box. This, of course is an obvious need and some one else has already done the hard work of implementing it, and quite well too, i must say: http://www.howjsay.com/.

Thursday, 13 February 2014

DataScience Indy meetup.

Topics suggested for the fishbowl:

1. Tuning models/parameters (SVM, NN, etc.,) [5]
2. Curse of dimensionality [5]
3. Spectral method [1]
4. Stories: DS implementations [5]
5. Temporal Methods/time series [1]
6. Cleaning data [5]
7. Cross-validation [2]
8. Stale data [2]
9. Billion dollar bracket [3]
10. Callable learning [1]

Numbers in [brackets] are the votes received; 2 per participant.

Topics 1,2,4,6 were discussed in the fishbowl.

Friday, 14 February 2014

You can be in one of these modes:

• Conscious execution of a task
• Putting out fires
• Passive consumption

It is the passive consumption you need to watch. You can often excuse away a lot of activity in the guise of “It is not wasted time like watching TV, so it is better”. But, there is only so much “unplanned preparation” you can do before it becomes a time sink.

Sunday, 16 February 2014

Experience happens when you keep showing up – something I can relate to after running DataScience Indy for an year. Many times, I worried that the topics were not interesting enough, there were not enough people attending the meetups, we had no sponsorships, etc., But the last meetup was a success and now I can look forward to a more energised group.

Monday, 17 February 2014

This comment summarises my experience at Tufte’s course quite well. (via)

I went to his course in 2005 as a figurative freshman, to use the analogy above. No experience with data visualization, despite working in Excel and Powerpoint almost full time. That day was one of the most eye opening experiences of my life. I suddenly felt as if I had been given a rule book, or a map, for all this stuff I thought was just made up.

I’ll admit to having joined the cult of Tufte for several years. I’ve developed a more nuanced appreciation for the strengths and weaknesses of his approach over time, but I think it’s important for people to realize the value he provides: Tufte is the Art History 101, the Intro to Economics, of data visualization. He’s giving you the fundamentals, the principles, the context in which to understand how visualization should work. For people who’ve been at this for years, it seems ancient and impractical. And it is. But just like art history, economics, philosophy, even computer science, one needs to understand what came before in order to innovate.

I read this site because of Edward Tufte. I think about how to present information in a clear and compelling way every day, in large part because of Edward Tufte. He’s medicine. Very worth taking, even if it isn’t sweet.

Sunday, 23 February 2014

• Sticking with a good idea

Sometimes, a good idea poses obstacles of its own. We should not let those obstacles beat us without a fight. Often, we just have to find a way to make it work.

Also, “High stakes programming”. Putting all your eggs in your basket is what’s needed sometime.

March

Saturday, 15 March 2014

When looking at a new programming language, if you keep looking for “maturity” of libraries, you will never get started in that language. You should be ready to write the libraries that you need for the immediate task at hand and see it maturity. Otherwise, you may as well stick to the established ecosystems, especially if you have to “ship it”. But remember, writing a library is probably the best way to level-up on the new programming language expertise.

April

Monday, 14 April 2014

ATS and OS development

I have been looking at ATS for a while for it’s correctness and ‘near-C’ level performance characteristics. It appears that writing an Operating System in ATS would be a good exercise to build a “provable”(?) system from the ground-up.

Here are some attempts to write a simple operating system:

• JamesM’s kernel development tutorials
• Kernel 101 Let’s write a Kernel - Arjun Sreedharan

Technical Presentation capturing

Two styles of capturing technical presentations have stood out for me the last couple of weeks.

1. The “picture-in-picture” style of presentation capturing the screen and the presenter. Seen here – Generators: The Final Frontier
2. Linking the video progress bar to the presentation PDF document so that the presentation changes page as the video progresses. This can be quite useful when following a code heavy talk, where one might want to copy-paste the code, URL etc., from the slide-deck. Seen here – Dependently Typed Programming in Idris: A Demo - David Raymond Christiansen

Knowing how to do both of these would be excellent for capturing the talks presented at our local meetups.

May

Sunday, 4 May 2014

Distributed Issue Tracking

The article about Github monoculture prompted me to think about a truly distributed DVCS (let’s call it TrueDDS - True distributed development system) that also has ticketing system, a wiki and other associated features necessary for collaborative development.

Update: May 7. This blogpost has a nice roundup of the current state of affairs as it pertains to distributed issue tracking.

Haskell-Beginners

Making much better progress on learning Haskell with Real World Haskell this time around. I ended up on #haskell-beginners after hearing it on twitter.

Was recommended on #haskell-beginners to use this set as a beginners guide to learn Haskell.

Tuesday, 06 May 2014

This is a good presentation explaining “Theorems for Free” paper.

Friday, 09 May 2014

Looks like it is time to leap frog into the front of the line of “Big Data” + “Data Science” by ignoring the entire “Hadoop” lineage of tools and directly going for Spark + MLLib. See:

• AMPCamp
• MLLib

Monday, 19 May 2014

Sometimes, a plain text file is the best way to maintain a calendar.

Structure:

Monday, 19 May 2014
- 10:00 - do foo
- 11:00 - do bar

Or may 19 stuff for the day may 19 8:00 do baz

What I liked about jac22’s calendar is that the calendar captures the activities right up to his retirement. That is something.

• Good research practices

Friday, 23 May 2014

From the preface of “Data analysis with Open source tools”:

Simple is better than complex. Cheap is better than expensive. Explicit is better than opaque. Purpose is more important than process. Insight is more important than precision. Understanding is more important than the technique. Think more, work less.

See also: The Zen of Python.

On reproducible research: You shouldn’t use a spreadsheet for important work

July

Monday, 7 July 2014

Scala Meetup

Brad Fritz - SBT

• Command line $ sbt "task1" "task2" unlike other command line tools eg: sbt "project dsl" "clean" "~ compile". Here,dsl` is the modulename.
• the tilde is for “watching” the project for changes.
• DependencyGraph – add to projects/plugins.sbt – addSbtPlugin("net.virtual-void" % "sbt-dependency-graph" % "0.7.4") and follow with sbt dependency-grpah to see the dependency graph.
• Global settings and Plugins:
  • $HOME/.sbt/0.13/*.sbt settings apply to all projects
  • $HOME/.sbt/0.13/plugins/ acts as global plugins project
• 0.13.5 introduces “auto plugins”.
• Pull Artifacts from Local Maven Repo use resolvers
• Resources:
  • @jsuereth - effective sbt.
  • Akka uses sbt well. Good source to study

Ross Baker - HTTP4S and SBT

• HTTP4S – WSGI for Scala.
• Blaze
• HTTP4S is multi-module, sbt reflects that.
• Dependecies are transitive.
• XML literals will be split off in 2.11.
• use Scala EStream over Iteritees.
• Unidoc plugin.- combined docs from all the modules of a project
• siteplugin -to generate website for the projects. also ghpages.
• Zinc Incremental compilation.

Sunday, 21 July 2014

• Love people, not things
• How to Be Happy - Less Wrong

Tuesday, 22 July 2014

Everything I Have Learned I Have Learned From Someone Else

Papers mentioned:

• Purely functional Data Structures by Okasaki
• Ideal Hash Trees by Phil Bagwell
  • implemented in Clojure
  • Very good complexity characteristics
  • RB Trees (?) paper is an update.
  • Hash-Array-Map Trie
• MiniKanren - PhD dissertation of Byrd.
• Efficient representation for triangular substitution: A comparison in MiniKanren by Bender, Byrd & Friedman.
• Equality for Prolog by Kornfeld
• How to make ad-hoc polymorphism less adhoc by Wadler
  • introduces typeclasses.
• Nominal logic programming by Cheney
• meaning of binding in lambda terms
• what unification means.
• Concepts, techniques and models of computer programming
  • constraint logic programming opens up prolog
• The art of the propagator by Radul and Sussman
  • related to constraint logic programming
• “finite domain logic/problem”
  • solvers - GECCO (sp?)
• Efficient multiple and predicate dispatching by Chambers and Chen
• efficient predicate dispatch..
• without the performance penalty of multi-methods
• Compiling pattern matching to good decision trees by Luc Maranget
• lazy pattern matching semantics for optimal decision tree.
• => less code generation.
• OCaml
• Views: A way for pattern matching to cohabit with data abstraction by Wadler
• matching on concrete types is a bad idea
• match on interface instead.
• Extensible pattern matching in an Extensible Language by Tobin-Hochstadt
• The art of of the metaobject protocol
• ” most important book on OOPS ” - Kay
• Compilation Strategies as objects by Anurag Mendhekar
• Communicating Sequential Processes by Tony Hoare.
• A multi-threaded higher-order user interface toolkit by Gansner and Reppy.
• used CSP to create a toolkit called eXene.
• Functional Reactive Animation by Elliott and Hudak

Functional programming is an important for interactive applications.

August

Thursday, 14 August 2014

From the profile of Nevanlinna Prize Winner Subhash Khot:

Khot had to figure out the answers to difficult problems by himself. Some problems took him six months to a year to solve. In the end, though, “I think that was very helpful that I learned everything the hard way,” he said. Now, Khot believes that his independence and ability to focus are his greatest strengths as a mathematician. “I’m perfectly happy to spend very long periods on a single problem,” he said.

Side note: Khot is from Ichalkaranji in Maharashtra, a place with Kannada name.

Wednesday, 20 August 2014

Programming Languages I care about:

• Python – the language i’m most familiar with
• OCaml – it is a practical functional PL. Fast, good books, pragmatic (fp with imperative hatches).
• F# – because it is like OCaml with fancy things like type providers, access to larger .NET framework.
• Purelang – dynamic with haskell flavour. feels useful for “programming in the small”.
• Rust – safe system level programming when it gets ready.
• Haskell – More FP?
• Erlang – build distributed systems without fuss.

Programming languages I do not care for, but want to have familiarity with:

• Golang – good tooling, popular for web, network stuff, community mojo.
• C – interop

Saturday, 06 September 2014

The world needs your excellence, not your perfection. src

September

Monday, 08 September 2014

MLI

Today, I came to know that Mark Irons passed away in Oct 2012. His “weblet” on Patterns for Personal Web Sites was a great inspiration for me in early 2000s. My reaction to the knowledge was that one of sadness. I did not know I could feel that for someone I’ve not even met, nor had I visited his site in a long time. In fact, I had to search my memory long for a keyword to look up his site. Then finally it came – “rdrop half”.

Here is a nice story of his life from a friend.

RIP MLI.

Wednesday, 10 September 2014

NOTE TO SELF: This is a log, do not treat it like a treatise and hold back from logging the mundane and the ordinary

NNTaleb is using André Miede’s ClassicThesis LaTeX style which is in-turn based on “The Elements of Typographic Style” by Robert Bringhurst. The ClassicThesis style looks very much like “Tufte style”. I wonder who influenced who.

In “Silent Risks”, NNT talks about Prof. David Freedman. Look into some his books “Statistics 92007)”, “Statistial models: theory and practice (2005)”.

Placed an order for “Statistics (4th ed)” by Freedman on Abebooks.com.

MB’s articles on typography are fun to read.

Thursday, 11 September 2014

Interesting read on “The Logic of Real and Complex Numbers” – via

webciten: A different take on saving web citations

A recent post by Pinboard.in brought up the issue of link rot. IIRC, more than 90%(?) of the links from 1997 are dead. There are many services including pinboard that offer to archive the bookmarks. What i’m talking here is slightly different.

We need a service that, say wimcite.org (Where Is My CITE . ORG)

1. Provide canonical citation URL like what Webcitations.org does ie., http://wimcite.org/abcd321
2. Store the snapshot of the webpage as the user saw it ie., a screenshot of the page.

The screenshot is important because, archival services see webpages as a document consisting of HTML + images + CSS files etc.,. The user does not care for all the artefacts, she cares to have a “proof” of having read that document and ready reference when it is required at a later date.

Features:

1. Use a bookmarklet to “wimcite” a webpage.
2. the user will be taken to the “wimcite” page ie., http://wimcite.org/abcd321 that will capture some metadata
   • URL of the page captured
   • time and date of capture
3. The “wimcite” page will embed the captured page as an image file (say .png).

An example of “Expository Systems writing” – A POSIX Queue Implementation.

Friday, 12 September 2014

• Like
• This
• ತಿಳೀತಾ?

Ordered a Raspberry Pi B+ and a case via Amazon.

Sunday, 14 September 2014

Learning self-control

sitzfleisch (Yiddish): the importance of continuing to work, regardless of the obstacles.

there are two warring parts of the brain: a hot part demanding immediate gratification (the limbic system), and a cool, goal-oriented part (the prefrontal cortex). The secret of self-control, he says, is to train the prefrontal cortex to kick in first.

To do this, use specific if-then plans, like “If it’s before noon, I won’t check email” or “If I feel angry, I will count backward from 10.” Done repeatedly, this buys a few seconds to at least consider your options. The point isn’t to be robotic and never eat chocolate mousse again. It’s to summon self-control when you want it, and be able to carry out long-term plans.

“We don’t need to be victims of our emotions,” Mr. Mischel says. “We have a prefrontal cortex that allows us to evaluate whether or not we like the emotions that are running us.”

Self-control alone doesn’t guarantee success. People also need a “burning goal” that gives them a reason to activate these skills, he says. His students all have the sitzfleisch to get into graduate school, but the best ones also have a burning question they want to answer in their work, sometimes stemming from their own lives.

Open Data Protocol

Dr Deming’s Management’s five deadly diseases

• Lack of consistency of purpoes.
• Emphasis on short term profits
• Annual rating of performance
• Mobility of Management
• Use of visible figures only

I have deleted by twitter account – https://twitter.com/btbytes today. On the whole, twitter is a time drain not a positive contribution to one’s growth. I must however

1. identify the benefits i got from using twitter
2. replace it with suitable alternatives that don’t have the same downsides.

One alternative for twitter is pinboard’s popular links page. Similar sentiments echoed.

Monday, 15 September 2014

Don’t multitask

heavy multi-taskers are worse at choosing which task to focus on. (“They are suckers for irrelevancy”, as Cliff Nass, one of the researchers put it.) Multi-taskers often think they are like gym rats, bulking up their ability to juggle tasks, when in fact they are like alcoholics, degrading their abilities through over-consumption. study

Wednesday, 17 September 2014

Read slowly to benefit your brain

WSJ Article

• Read like a first grader
• Go to a place with no computer, laptop, iphone etc.,
• Take occassional notes. It’ll help you think about your reaction to the book
• Treat reading as an exercise.. something that you make time for.
• dedicate at least 30-45 minutes to reading. Anything less won’t allow you to immerse yourself in it.

I arrived at the same realisation yesterday after cutting down on indiscriminate internet usage, avoiding unnecessary “screen time” and focussing on reading a fairly technical paper (by Lamport).

I wish all researchers follow Leslie Lamport’s manner of documenting their publications. Piecing together the flow of ideas is often the most difficult thing for an autodidact.

I came across S M Ulam in an online forum. To revisit at a later time. MAA profile.

Prof Larry Wasserman’s now defunkt blog – Normal Deviate. Another candiate for “blog to compiled book aka BlogBook” project?

I recently purchased the late Prof Freedman’s “Statistics” 4th edition. The book is very readable and has many exercises. This made me search for his other books because I remember both NNT and Shalizi talking about his work. I found Statistical Models: Theory and Practice . Interestingly, NNT had glowing reviews of the book. Saved for later purchase.

Came across “rational-choice theory” twice today, in two very different contexts.

In 1968, one of the founders of rational-choice theory, Gary Becker, wrote a pioneering paper, “Crime and Punishment: An Economic Approach”.

George Mason University, a bastion of rational-choice theory.

Peter Leeson wrote The Invisible Hook, a 2009 account of the economics of piracy.

Skarbek’s doctoral adviser, Peter Boettke, showed how the behavior of the Soviet economy actually made sense if you viewed it as controlled not just by the government but also by the black- and gray-market activities of citizens.

And, in review of “SM:TP”.

NOTE: You cannot jump into the middle of a mathematics paper or a book and expect it to “make sense”. Give yourself a chance!

Artificial sweeteners induce glucose intolerance by altering the gut microbiota

Non-caloric artificial sweeteners (NAS) are among the most widely used food additives worldwide, regularly consumed by lean and obese individuals alike. NAS consumption is considered safe and beneficial owing to their low caloric content, yet supporting scientific data remain sparse and controversial. Here we demonstrate that consumption of commonly used NAS formulations drives the development of glucose intolerance through induction of compositional and functional alterations to the intestinal microbiota. These NAS-mediated deleterious metabolic effects are abrogated by antibiotic treatment, and are fully transferrable to germ-free mice upon faecal transplantation of microbiota configurations from NAS-consuming mice, or of microbiota anaerobically incubated in the presence of NAS. We identify NAS-altered microbial metabolic pathways that are linked to host susceptibility to metabolic disease, and demonstrate similar NAS-induced dysbiosis and glucose intolerance in healthy human subjects. Collectively, our results link NAS consumption, dysbiosis and metabolic abnormalities, thereby calling for a reassessment of massive NAS usage.

THINK: Atomic directoy moves. Replace directory A with directory B atomically. If the move fails for whatever reason, restore A.

If you are serious about writing, you must have your own document processor

There are a lot of document processors, including my favourite - pandoc. Every new release improves on the state-of-art and adds more features. Perhaps pandoc among them is unique in the regard that it tries to capture the semantics of the document (what it means vs presentation) that harkens back to HTML, TeX etc., However, for the sake of keeping the format simple and universal, the programmer of such a system keeps the feature set large enough to satisfy most users’ common need, yet general enough. If I were to borrow a programming language analogy, “don’t do a FORTH”. However, what is good for the maintainer of a popular program may not be true for the individual user. The individual has a unique, varying needs and they change over time. It is often tiresome to fall back to using the same common-denominator features. So, an enlightened user must attempt to grow a toolchain over the substrate of a popular document processing toolkit, in my case it might as well be pandoc.

Tuesday, 23 September 2014

Large scale study says group nature walks linked with significantly lower depression and perceived stress, enhanced mental well-being

From the reddit discussion:

I found interesting historical evidence from World War II suggesting that neurotics living in London found their average level of anxiety actually declined during the Blitz–they had real things to worry about (bombs dropping on their heads) and also they felt relieved to see other people outwardly expressing the anxiety the neurotics felt all the time. Also, some sociologists have suggested that the Middle Ages were low in anxiety because a) they were so high in real danger (disease, murder, etc) and b) there was so little free choice, which actually reduces anxiety.

Life in the Middle Ages was grim and awful–but may have been not particularly high in anxiety of the sort we denizens of the post-industrial capitalist age are – Scott Stossel

Sunday, 28 September 2014

Looks like yahoo is shutting down “Yahoo Directory”. Yahoo directory was the first “Internet” I used. The Infantry Road(?) NIIT had put out a newspaper ad about “1/2 hour of free internet”. Guru and I went to check it out. The first screen was “Yahoo directories”. They gave some speil about NIIT training afterwards. End of an era.

October

Wednesday, 01 October 2014

Two important things I remember from (still reading) “Chandra”:

1. Make time for all the things you want to do by having clear seperation. (Chandrasekhar used to have a seperate office for his Editor’s role and strictly enforced the office times even with friends and colleagues).

2. “Systamization” Answer this question: “What am I going to achieve with this new endeavour? How long is it going to take?” Once you have the answers to this, complete that phase and move on to the next one. Plan the closure well in advance.

Saturday, 04 October 2014

I wish more professors would publish their lectures and materials like Sanjoy Mahajan.

He is an expert user of ConTeXt typesetting software.

2014-11-25: The art of insight in science and engineering by Mahajan.

Wednesday, 08 October 2014

“knowledge which comes without any effort from the student is completely useless for that student. He believed that true knowledge comes from personal experiences and individual confrontations actualized by one’s own intentions.” – Beelzebub’s Tales to His Grandson

Friday, 10 October 2014

Different types of Journal Paper Types. These are useful distinctions to know.

Tuesday, 28 October 2014

William Feather once wrote that being educated means “being able to differentiate between what you know and what you don’t.”

In an article by Dunning (of Dunning-Kruger effect)

In between listening to Taleb on “technology”, (nerdery) and reading this interview, how does one continue to work on technology? One thought to keep in mind is: “Your work should reduce the impact of technology so that it takes away all the make-believe work it has created for us. If your technology can take away the complexities technology has introduced into people’s life and give the time back to them, it will be a contribution.”

Wednesday, 17 December 2014

With the start of dedicated the assorted links, I see this becoming more of a log (blog) than a quasi-link-aggregator that it had become over the last few months.