Lab Notebook

(Introduction)

Coding

  • cboettig pushed to master at cboettig/labnotebook: toggle off pageviews since cannot deploy via travis 08:26 2014/04/15
  • cboettig commented on issue ropensci/reproducibility-guide#43: I think this thread highlights some pretty interesting issues visa vi reproducibility, collaboration and technical barriers. On one hand, I agree w… 08:22 2014/04/15
  • cboettig opened issue Sija/garb#33: Unable to authenticate from Travis 08:08 2014/04/15
  • cboettig pushed to master at cboettig/labnotebook: updated semantic data. also asked google analytics to allow login update semantic data 06:55 2014/04/15
  • cboettig closed issue cboettig/labnotebook#98: tags link to their research project page? 06:35 2014/04/15

Discussing

Reading

  • The limitations of diversity metrics in directing global marine conservation: Marine Policy (2014). Pages: 123-125. James P.W. Robinson, Easton R. White, Logan D. Wiwchar, Danielle C. Claar, Justin P. Suraci, Julia K. Baum et al. 05:26 2014/04/11
  • Integrating abundance and functional traits reveals new global hotspots of fish diversity.: Nature (2013). Volume: 501, Issue: 7468. Pages: 539-42. Rick D Stuart-Smith, Amanda E Bates, Jonathan S Lefcheck, J Emmett Duffy, Susan C Baker, Russell J Thomson, Jemina F Stuart-Smith, Nicole a Hill, Stuart J Kininmonth, Laura Airoldi, Mikel a Becerro, Stuart J Campbell, Terence P Dawson, Sergio a Navarrete, German a Soler, Elisabeth M a Strain, Trevor J Willis, Graham J Edgar et al. 05:26 2014/04/11
  • A bayesian approach to identifying and compensating for model misspecification in population models.: Ecology (2014). Volume: 95, Issue: 2. Pages: 329-41. James T Thorson, Kotaro Ono, Stephan B Munch et al. 06:42 2014/04/05
  • Quantitative Bioscience for the 21st Century: BioScience (2005). Volume: 55, Issue: 6. Pages: 511. Alan Hastings, Peter Arzberger, Ben Bolker, Scott Collins, Anthony R. Ives, Norman a. Johnson, Margaret a. Palmer et al. 05:23 2014/03/28

Entries

Notes

11 Apr 2014

labnotebook

Updates to configuration to be more portable and lightweight:

  • Drop APIs for Mendeley and Github in favor of just pulling from atom/rss feeds. Doesn’t require a platform-specific API or authentication, and has proven more robust, if sometimes less flexible and fancy.
  • Drop git-modified and sha timestamps. It’s nice having these on the pages and included in the rdfa, but the information is accessible from the git repository and from Github. The overhead time in computing these values and successfully including them wasn’t worth it.
  • However, add a sha for the vita page using a simple filter.
  • Wishing I could just embed svgs into md output as XML, instead of as binary. See full description of issue: yihui/knitr#754.

  • Installed vim-pandoc’s dev versions, vim-pantondoc and vim-pandoc-syntax via pathogen; seem to be working rather nicely.

reproducible paper example

from richfitz mwpennel and co: A manuscript with examples executed by makefile (or R source script) and checked and built by travis: richfitz/wood. Wow. Particularly nice jobs in:

  • handling downloading of raw data from original (Dryad-based) sources. Makefile seems to handle doing this in a cache-friendly way that is lighter on Travis network.
  • providing the cleaned, process data as seperate data files in clearly marked output directory. Still, the output would be more modular if done in some more generic format than R’s binary serialization as rds files. They do also provide data on Dryad, which is presumably the processed data in a more sensible format, though their data DOI, 10.5061/dryad.v7m14, is still embargoed in advance of publication at the time of writing.
  • Travis checks not only that code runs, but that it downloads and that the manuscript and figures can be complied (by LaTeX), with supplement knitted and compiled as html.
  • in a very nice touch, Travis pushes html and pdf results back to Github on gh-pages.

Some very minor nitpicks:

  • manuscript isn’t strictly a literate (knitr) document, though figures are being produced on the fly from code by make. May be a sensible choice (though misses a few opportunities for dynamic inline values).
  • license not declared, but already coming soon richfitz/wood#8
  • discussion of citation of code: richfitz/wood#11
  • trouble in executing locally, see richfitz/wood#12

rfishbase

Read more



Multiple Uncertainty Notes

10 Apr 2014

multiple uncertainty

  • Set up Paul Fackler’s MDPSOLVE and run Paul’s code. Installing user-contributed extensions in matlab seems mostly a matter adding the unzipped source directories to the path, e.g. as so:
currentdir = cd
cd /home/cboettig/.matlab/MDPSOLVE
addpath(genpath(cd))

cd /home/cboettig/.matlab/plot2svg
addpath(genpath(cd))

cd(currentdir)

(I still find matlab/octave’s syntax of not distinguishing character strings from variable names from function names kind of terrifying, but I guess that’s the price of working with economists.)

Code runs, with svg output using the user-contributed svg function plot2svg since Matlab doesn’t seem to include such basic functionality (though works fine on octave).

ropensci

Morning mostly spent wrapping up proposal steps

  • Pandoc-latex formatting stuff (xelatex, Times New Roman 12pt, doublespacing, makefile configuration, latex-template configuration, margins, yaml header metadata).

  • Reworking introduction into seperate sections

  • feedback on further issues through issue tracker.

Reading

  • Further evidence that anything can happen with regards to transitions in spatial dynamics: Gowda et al, Phys Rev E

Read more



Multiple Uncertainty Notes

09 Apr 2014

Attempt to replicate Figure 3, but with one noise at a time

  • Code: one_noise_at_a_time.m
  • Uniform Noise
  • Large noise means devation of 0.5, as in Sethi
  • Other noise terms are 0 instead of 0.1

Figure 3

  • Code: carl_fig3.m
  • Uniform Noise
  • Large noise means devation of 0.5, as in Sethi
  • Small noise means 0.1, as in Sethi

One noise at a time, lognormal noise, sigma = 0.3. (code)

Read more



Multiple Uncertainty Notes

08 Apr 2014

Growth Noise Only

Run:

matlab -nodesktop < testing.m > testing.log

Results from running testing.m:

  • log normal noise
  • sigma_g = 0.2
  • Other noise set to zero.
  • Coarse grid 0:5:150
  • (See linked code all parameters)
Growth noise only
Growth noise only

measurement Noise only

measurement noise only
measurement noise only

Implementation Noise only

Implementation noise only
Implementation noise only

  • Thinking about the best way to embed SVGs in the notebook, see docs/21
  • Thoughts on date metadata, see docs/20

Read more



Notes

07 Apr 2014

ropensci

Working on proposal. Discussion of section 2: other players in the domain (or rather, the many community interfaces with rOpenSci).

other

rfishbase

Add FAO areas (from 04/05), see rfishbase/20

EML

  • Discussion of unit handling, feedback from Karen. EML/12

  • Need to update creation of filenames for csv files following Matt’s advice: EML/106

  • No luck with XMLSchema::readSchema parsing the measurement unit schema definitions in STMML

> XMLSchema::readSchema("../stmml.xsd")
Error in FUN(X[[1L]], ...) :
  formal argument "localElements" matched by multiple actual arguments

misc/reading

  • teary: discuss interviews. send copy of materials for archive.
  • review request EcoLet
  • follow-up emails from Wisconsin visit. Follow-up reading:

  • RDCEP

  • Here’s the RDCEP robust control paper

  • Here are two papers on the value of experimentation, 1, 2. Key references are the two papers by V. Wieland cited there. Steve and Buz discussed one of these in “Panceas” paper

  • Here’s a link to the BDW Brookings paper which has an interesting discussion by Sargent and Sims

  • This one proposes an approach to reporting to policy makers that respects uncertainties

  • Here is one we did for the World Bank to deal with uncertainties in growth policy

  • Since optimal control and even robust control can lead to controls that are complicated functions of the state vector, there’s a literature in applications to central banking that stresses “simple rules”, here’s a recent example from our shop that has lots of references,

  • Finally, on EWS’s, there’s a general multivariate mathematical approach to variance-convariance matrix approaches to EWS’s in the Mathematical Appendix to this paper by Biggs et al.

Read more



Wisconsin Visit

03 Apr 2014

Carl Boettiger Schedule 3-4 April

Thursday

Person/activity Time Location
Arrive 3:00 Dane County Airport
Tony 4:00 pickup at hotel
Jesse Miller 4:30 Birge 448
Kyle Webert 5:00 Birge 457
Cristina Herren 5:30 Birge 456
Dinner with Tony, Karen Strier 6:00 leave from Birge Hall

Friday

Person/activity Time Location
Breakfast, Meghan Fitzgerald and Fan Huan 7:30 pickup at hotel
Cecile Ane, Lam Si Tung Ho 9:00 Birge 341
Steve Carpenter and CFL folk 10:45 CFL 226A
Jeremy Ash 2:00 Birge 338
Break, walk over with Jacob 2:30 Birge 460
prep for seminar 3:00 Nolan Hall
seminar 3:30 Nolan Hall
reception 6:00 Tony’s house

Misc notes:

  • See email from Buz for robust control reading. email to Karen.

Read more



Ropenhack Notes

31 Mar 2014

Breakfast

  • discussion of what makes an API good or a total pain
  • discussion of extending data.set class with metadata

Morning Session

Chatting with Hadley:

Best practices for R client packages to REST APIs

  • (When) Do we return response call metadata?
  • Very modular function construction
  • Error handling, assertthat checks
  • Explicit parsing, as seperate function. e.g.
  • Better alternatives in place of require when loading SUGGESTS lists, which still pollutes the namespace; see examples in httr.

We’ve begun writing dependencies package to address some of the issues in managing version dependencies of R packages.

Afternoon Session

Working with Martin and Gavin on docs workflow.

One challenge of this setup is passing secure data over travis, such as the keys necessary to execute the git push. Travis has a mechanism for encrypting and transmitting these keys.

Otherwise, working in simplifying and sanitizing the separation between ruby and knitr interfaces in providing a dynamic coding static blogging platform.

Also various interactions and video interview for event documentation.

Evening

Excellent social; discussions with Josh Bloom re BIDS etc

Wednesday

Mostly hacking on docs workflow, see commit log etc.

Read more



Review Of Progress On Packages

24 Mar 2014

A lot of the past week has been focused on some catching up and tidying up in my active projects in metadata manipulation and publishing tools, EML, RNeXML, and rfigshare. The full story of the week and half unfolds on github, but summarizing progress here.

rfigshare

Changes to httr interface orginally broke rfigshare, though working with @hadley the original mechanism was restored through sign_oauth1.0(). The changes provided a much nicer mechanism for handling oauth authentication without having to specify tokens directly, and resulted in a relatively straightforward overhaul of the authentication mechanism in the package, while still preserving the package API. During this process I also tackled a bunch of outstanding issues, such as: pretty-printing results, fixing behavior of search and adding authors, getting embeddable image URLs and flushing out the test suite and activiting full travis integration of the tests. See package NEWS for a more complete list (v0.2-9 and v0.3). This version is now on CRAN.

In the process I learned a bit more about httr’s oauth handling.

RNeXML

Main focus has been moving manuscript text foward, flushing out a more complete description of package functions, organization, and use cases. In particular this has focused on:

  • Outlining the tiered levels of metadata manipulation (from the elementrary get_metadata() to the XPath and SPARQL queries) and metadata generation (from title and creator to automatic taxonomic reference links using taxize to building on existing and custom ontologies; and the related example of extending the schema itself using metadata (simmap).

  • Finalizing and writing out this simmap example was the other goal, illustrating extending the package.

Lost too much time on some of the fine points of travis integration with package dependencies on specific users and branches of Github, on omegahat, and cases where apt-get cannot get background dependencies automatically.

EML

Still need to continue mansucript writing and extending functionality; meanwhile minor changes to reflect updates to rfigshare and to travis integration.

rfishbase

Some momentum on rfishbase2.0 thanks to involvement from @tpoi

Read more



Notes

18 Mar 2014

ropensci

pdg-control / policycosts

  • Started tidying up this repo’s manuscript folder a bit.
  • Merge Paul’s various changes on tex and Word documents into Rmd

Misc

  • Whoops. Couldn’r resist a wee rant on G+, 2014-03-02 on this #PLOSfail discussion.

Read more