Here is the latest bag of tweets^*, which covers February 2012.

(*) These are interesting news that I found on Twitter and that I archive periodically.

• siah: Reading “Fast SVD for Large-Scale Matrices” highly recommended http://t.co/REMMndoH (25 Feb)
• StatFact: “The persistent topology of data” http://t.co/5dSmhkeI // applying computational algebraic topology to data analysis (24 Feb)
• prometheia: “Nonlinear Principal Components and Long-Run Implications of Multivariate Diffusions.” By Chen-Larsen-Scheinkman http://t.co/dAY06a98 (24 Feb)
• Biff_Bruise: RT @OECD_Stat: Try the future new version of OECD.Stat in beta with an improved user interface and dynamic graphs: http://t.co/Fgfc3cHG (24 Feb)
• FGRibreau: Learn from Haskell - Functional, reusable JavaScript http://t.co/QRgbMOiH (24 Feb)
• tomstafford: Received 14 June 1995; accepted 8 May 2006 Available online 8 October 2008 Academic publishing is fantastically broken http://t.co/e3ZmVipp (24 Feb)
• mbostock: @abhay There’s a nice list of bitwise hacks here: http://t.co/cGywqIbK (24 Feb)
• StatFact: Prediction: the Lasso vs. just using the top 10 predictors http://t.co/XMMwz0Jx (23 Feb)
• ijuma: “Fast number crunching and fast time to market with Scala” http://t.co/qV0ZZQZX http://t.co/mStyLfcH (23 Feb)
• albert_swart: Important discussion on false positives, model selection and overfitting in ML/NLP https://t.co/crPleJ6b via @prismatic (23 Feb)
• rob_aldridge: Interesting paper examining whether highly tweeted articles receive more citations http://t.co/jGpy0i4x (23 Feb)
• sc_k: The case for open computer programs http://t.co/EGCKi9Y3 (22 Feb)
• tkb: The Open Data Handbook is out! Civil servants, journalists, activists, developers, researchers - take a peek: http://t.co/4rTZbZyW #opendata (22 Feb)
• UnixToolTip: Better bash history http://t.co/XNFqVP1t (22 Feb)
• leed0: interesting scripting language comparison table: http://t.co/EGu0zRO4 (22 Feb)
• daringfireball: Mountain Lion Details: http://t.co/kxCdEh45 (22 Feb)
• neilfws: polar histogram: pretty and useful http://t.co/DnQmuiAq (21 Feb)
• JanWillemTulp: Recap of visual.ly’s #d3js meetup, including slides and video http://t.co/nERLWT0g (21 Feb)
• sushantsinha: Python code for pdf to text conversion. Tries pdftotext first and then calls tesseract/ocropus. Hides img format issues http://t.co/ZRdShsKA (21 Feb)
• kdnuggets: Peter Norvig (Google Dir of Research): The Unreasonable Effectiveness of Data - must see for all #BigData miners http://t.co/6EN0Kedk (21 Feb)
• lizardbill: Visualising Stack Overflow Tag Relationships http://t.co/4l9IV0zj (21 Feb)
• mikedewar: Finally got all the examples in the new remake of #d3py working. Only x-axes so far but they’re already looking pretty! http://t.co/RIbSh9rr (21 Feb)
• siah: Online Python Tutor http://t.co/K9R82KrH (21 Feb)
• berkeleymalagon: Cool read! A history of O’Reilly book animals. http://t.co/MwBmNkC3 #tech #books #oreilly (20 Feb)
• undunc: Excellent (downloadable) book on risk by Brit Med Assoc http://t.co/wNhf1Pvw For docs, but fine for everyone (20 Feb)
• ChrisDiehl: Why We Created Julia http://t.co/z2RAhrkt via @prismatic (20 Feb)
• berndweiss: Shige’s Research Blog is one of the best one-sentence blogs I’ve ever seen http://t.co/f7128fPh #latex #rstats #stats #linux (20 Feb)
• emacs_knight: From the blog - “Package Management in Emacs: The Good, the Bad and the Ugly” http://t.co/x7gPGS56 (20 Feb)
• emacs_knight: From the blog - “Color Theming in Emacs: Reloaded” http://t.co/JcRNP5vQ (20 Feb)
• recology_: Problem 4 #rstats on OSX? @HackerNewsYC: Apple Removes X11 in OS X Mountain Lion, Shifts Support to Open Source XQuartz http://t.co/kEi4Mmdk (20 Feb)
• debasishg: The Architecture of Open Source Applications: Berkeley DB chap is a must read http://t.co/K8FQrXly via @prismatic (19 Feb)
• MonkScott: Results from a Semi-Supervised Feature Learning Competition: http://t.co/P53Hb0eP #machinelearning (18 Feb)
• ramhiser: Flabbergasted! Ripley said “I am sorry” on #rstats help. https://t.co/IUnYgOUZ (18 Feb)
• neilfws: Sweave for a book - A Sweave Wiki http://t.co/WQTyShWn #diigo (18 Feb)
• TeXtip: How to create a crossword puzzle in LaTeX http://t.co/kTe8EiW4 (18 Feb)
• MathematicaTip: Usually you should increase WorkingPrecision and PrecisionGoal together: NIntegrate[x^x, {x, 0, 1}, PrecisionGoal->50, WorkingPrecision->60] (17 Feb)
• drewconway: Many people asking for the code to go with Machine Learning for Hackers, @johnmyleswhite is hosting on his Github http://t.co/HX4OpREa (17 Feb)
• sithburns: Apple Unveils Mac OS X Mountain Lion: So Long iChat, Hello Twitter [PREVIEW] http://t.co/dcZs9ufk (17 Feb)
• BMJ_Open: Low self-rated health associated with increased mortality in those with #diabetes http://t.co/GIQrdZZE (17 Feb)
• moclanmomo: Wolfram|Alpha Pro wants to democratize data analysis http://t.co/Nfdky66u As long as they don’t make it #FOSS, that is just marketing. (17 Feb)
• wellcometrust: Computer programs may be able to identify individuals most at risk of anxiety and mood disorders. http://t.co/xzBiCVTZ (17 Feb)
• druvus: How Identical are Identical Twins at sequence level? http://t.co/uh63mipr #sequencing #snp (17 Feb)
• DataMiningTips: Weave, a “ridiculously powerful” open-source data visualization http://t.co/sNj35pwm #kdnuggets (17 Feb)
• Rbloggers: Generating Twitter Wordclouds in R (Prompted by an Open Learning Blogpost): (This article was first pub… http://t.co/9QAUEwCr #rstats (17 Feb)
• mathematicsprof: Leading statistician (John Marden) puts his excellent book on multivariable statistics online in pdf format. http://t.co/rlTCUcdb (17 Feb)
• statsepi: What #epidemiologists do. http://t.co/XzxVTjCa @Peter_Tennant @gingerly_onward @tinscience (16 Feb)
• Elenina_M: NumPy / SciPy related tutorials at PyCon 2012 #pycon #python - http://t.co/CkJqtRyn (15 Feb)
• signmagazine: Review: Understanding Measurement: Reliability by Patrick Meyer http://t.co/c0ipS7sK (14 Feb)
• flowingdata: Interactive and animated word cloud http://t.co/ZNZ3tLbP (14 Feb)
• jseabold: Making Grapical Models with PyDot (or why I’m cooking up a matplotlib solution) http://t.co/aogAvyUl [Blog post] (14 Feb)
• JanWillemTulp: I wonder how much #Papaya, the #statistics lib for #Processing http://t.co/UKl52nnN differs from #Apache Commons Math http://t.co/5YfUWGIi> (13 Feb)
• stevengustafson: A list of 300 major algorithms http://t.co/WUIjnj5G from different domains #compsci #kd (12 Feb)
• mbostock: Force-directed Voronoi/Delaunay. Use mousewheel to add/remove nodes. Wow! http://t.co/QYAtKKTQ by @cmanning88 (12 Feb)
• vsbuffalo: + <- function(x, y) .Primitive(’+’)(.Primitive(’+’)(y, x), runif(1)) put that in your ~/.Rprofile. #lisplanguagelove (12 Feb)
• GeneticsUpdate: Genetic risks for type 2 diabetes span multiple ethnicities http://t.co/XpwkyLBN (11 Feb)
• druvus: Python Data Analysis Library http://t.co/5KYHmXGE #r (11 Feb)
• algoriffic: I agree wholeheartedly: Mathematica: A Love Story http://t.co/ccrvdsak via @prismatic (11 Feb)
• marksim: Just released the-much improved- 0.5.0 version of Omelettes - your database scrambler (for obfuscating sensitive data) https://t.co/yC21PV1G (11 Feb)
• SmartTypes: only data with still unknown but learnable statistical or algorithmic regularities is truly interesting – http://t.co/2c4Q44vM (11 Feb)
• TeXtip: Why some people believe you should not use eqnarray http://t.co/8Efp8LRt (10 Feb)
• JanWillemTulp: nice #dataviz work by @Periscopic http://t.co/Gm6dL3Tr nice touch that sound is being used again in the visualization! (10 Feb)
• gd047: A Backstage Tour of #ggplot2 with Hadley Wickham. Presented: Wednesday, February 8, 2012 #rstats http://t.co/mj2gemrL (10 Feb)
• planetpython: John Cook: Mixing R, Python, and Perl in 14 lines of code http://t.co/0zeEyge1 (10 Feb)
• SoutenonsLeMur: #Autism in #France: Support Statement Prof Anita Thapar, Clinical Professor of Child & Adolescent #Psychiatry, #Cardiff http://t.co/q6IcR0W1 (9 Feb)
• alexablag: Dear useRs, pls check out the package for reproducible report templates that @daroczig and I developed http://t.co/I5OLimTk #rstats (8 Feb)
• dgmacarthur: “…publishers need academics more than academics need publishers” - great @economist piece: http://t.co/b0QAnWoG (7 Feb)
• robertlayton: Bleeding edge #scikit-learn now has a parallel version of k-means. Pass a value to the n_jobs parameter to use multiple processors. #sklearn (7 Feb)
• ChrisDiehl: Numeric Javascript, supporting numerical computations in JS http://t.co/siUqVXVT via @prismatic (6 Feb)
• lizardbill: #SICP 2.52: Levels of Language for Robust Design http://t.co/e5elNEaM (6 Feb)
• tomstafford: Is anyone doing the sci-art of this? They should: Visual Image Reconstruction from Human Brain Activity http://t.co/2pzA1n2A (4 Feb)
• kdnuggets: the Art of R Programming: lots of gems, including parallel R http://t.co/N5Fvqadg A Valentine gift for data scientists #datamining (4 Feb)
• GaelVaroquaux: Label propagation for semi-supervised learning in @scikit_learn by Clay Woolam. Beautiful work https://t.co/Uy9XnfR0 (2 Feb)
• mja: Awesome looking short course in quantitative population genetics in Edinburgh this summer: http://t.co/HZQxZ4g3 (2 Feb)
• _onionesque: Interesting Manifesto: Machine learning is a cottage industry, not an engineering discipline. […] http://t.co/Hc6UMdWW (1 Feb)
• nytimeshealth: Op-Ed Contributor: Asperger’s History of Over-Diagnosis http://t.co/DwqlJVbU (1 Feb)
• jrauser: Amazing visualizations of the molecular machines operating in your body right now: http://t.co/xduyxt5H (1 Feb)