Lab Notes
This section includes links to various notes I've put together as part of my work. They are organized by date, and include a short description. Feel free to contact me with any questions or comments you have about them, or to reuse or redistribute the material in ways that are helpful.
4 September 2024: Some musings on trying to estimate observation error rates for a case where error is correlated with system state: <link>
4 April 2024: Photos for new Biodiversity/Ecology Masters Program at Uni Graz: <link>
6 March 2024: Simple example of biases arising in random effects models due to model misspecification. <link>
22 September 2023: Some musings on grant applications and collaboration, following funding of my first "big" research project as co-PI. <link>
22 March 2023: An attempt at a locally linear (ODE) metacommunity model. Sadly gets bogged down in a somewhat too complex spatial moment model. <link>
9 October 2022: R script for changing the order of elements in file names based on a separator (e.g. "_"). <link>
4 September 2022: LaTeX code used for making my CV. <link>
2 September 2022: Example script for estimating the Lyapunov exponent of a process using the rEDM package. <code>
8 January 2022: Notes on COVID growth rates in the Omicron surge following super-exponential growth. <link> <code> <data>. See Chapter 5.4 in the Lehman et al. eBook for details on the model <link>.
3 June 2021: Very simple code for measuring light intensity with an Arduino, adapted from James Carlson. <link>
8 April 2021: An example showing when random effects that overlap perfectly with a categorical variable can lead to biased estimates (at least for lme4, the result is that the fixed effect is correctly estimated, but the random effect is set to zero). <link>
23 March 2021: A short discussion on how to measure the rates of an exponential process, either in terms of averages or half-lifes (and why half-lives are better). <link>
7 March 2021: A basic SIR model with re-infection. <link>
13 February 2021: A very simple power analysis comparing results from the recent study on AstraZeneca vaccine efficacy against the B.1.351 "South African" COVID variant. <link>
3 February 2021: An extended example for power analysis, showing how to deal with multivariate predictors. <link>
28 July 2020: Transforming Lotka-Volterra parameters from the linearized to the classic formulation - shows a short example with two species. <link>
11 May 2020: Weird things that R does - A compilation of notes related to R's many irrationalities that can stymie people who are just starting out with it, and some of my solutions for these problems. <link>
19 April 2020: A comparison of fitting methods for projecting a logistic growth curve, inspired by this post. Interestingly, the simple rate based fitting method performs much worse than a state-based method, which I would not have thought. <link>
21 March 2020: A more polished version of the SIR model, for making short-term forecasts in a few countries. Please see disclaimers in the file. <link>
10 March 2020: Update, testing per-capita growth rate of COVID-19 vs. number of cases. Requires data from the CSSEGISandData 'COVID-19' GitHub, available at https://github.com/CSSEGISandData/COVID-19. <link>
24 February 2020: A basic SIR model, based on rough current data about COVID-19. Note that this model assumes no changes in behavior or uncertainty in rate estimates, and is therefore highly unrealistic (i.e. don't take the forecast too seriously). <link>
9 January 2020: Some basic notes on applying power analyses with binary and continuous data. <link>
14 November 2019: Some example code and data for analyzing human population growth, and the competition/predation data from Gause 1934. For use as part of the Fall 2019 ecology course at Halle. <zip>
17 September 2019: Some notes on coexistence and stability that I've assembled as part of literature reviews for a few projects that I am working on. <pdf>
2 September 2019: Somewhat trivial simulation showing expected uptake of light by individual species in a multi-species mixed canopy. <R>
15 July 2019: Example analysis in response to a recent paper on potential logit regression, tweeted by Forian Hartig (thread available here). <R>