User:Ben Moore/figures

This page shows the methods and data I've used to create figures uploaded to wikimedia commons for use in articles. The idea is to make it as easy as possible for anyone else to replace the figures I've made with better ones! The code provided is in the R programming language, and I recommend the great open source IDE RStudio for playing about with it. Let me know on my talk page if you have any questions or criticisms.

Data
Two data sets offered by the US gov (.xls excel spreadsheets):


 * Unemployment rate
 * Monthly net change in employment

NB. These seem to be kept reasonably up to date, so these plots can be updated as new data becomes available.

Data
Table from NHGRI, should be updated "frequently".

Comments
The source code is no longer accurate, as I updated and redid the data using gnuplot. (It seemed easier for simple cases. Now that I'm learning R, that might not be the case further on, but regardless, the code on the image page doesn't match what's above. grendel|khan 05:51, 27 July 2016 (UTC)

Data
A  containing the growth information up to (an extrapolated) month in late 2013 can be downloaded here from the NCBI website. The sequencing platform data is hardcoded below but was retreived from this publication.

Comments

 * The dark green line could be raised above the light green
 * Pie charts are 'bad'
 * Yellow = other?
 * This pie-chart isn't particularly bad. There are only 4 categories. I don't think a barchart would be much better for illustrating the point.
 * Thanks for the comments Ppgardne, I know you're a bit of a graph pro! Yeah yellow is other but I couldn't find any combination of parameters that would include this label, not run over the others and still remain a readable size… There's probably some adj or offset I've overlooked—haven't played with pie charts much before  Jebus989 ✰ 22:28, 13 January 2013 (UTC)
 * a simple "text(x,y,'Other')" might do the trick. No tick mark though. Or refine the plot in inkscape -- that should keep all your SVG loveliness. --Paul (talk) 00:11, 14 January 2013 (UTC)
 * BTW, that "pars=(cex=0.6)" is not doing what you think it is. Remove it and do a "par(cex=0.6)" before "subplot" instead. --Paul (talk) 03:56, 16 January 2013 (UTC)
 * You're right cheers for picking that up. When the labels are actually scaled down they are too tiny, especially for a thumbnail. I reckon using text is a good option, maybe with a custom tickmark via segments if I find myself with time to waste—I wanna try keep it reproducible if I can so no Inkscape for now  Jebus989 ✰ 22:34, 16 January 2013 (UTC)
 * Thanks for sharing this. I've already used it for a few projects. What if you scale "cex" down and increase "size"? You have heaps of space and 1.5 is a little small. --Paul (talk) 04:33, 17 January 2013 (UTC)
 * ✅ Good idea! Now all segments labelled, cheers. And glad it came in use to someone!  Jebus989 ✰ 10:53, 17 January 2013 (UTC)
 * Redraw with ggplot2?

Data
The size of EMBL-Bank in nucleotides is reported in each set of Release Notes. This graph was drawn from Release Notes for version 114 (currently available here as of Jan 2013). For use in the code below, the relevant table was copied from the report and pasted to a file ( below).

Comments

 * The two lines probably don't need to be on different axes
 * Minor tickmarks highlighting the log scale might look nice
 * Points of interest could be labelled; i.e. Human Genome Project; milestones
 * coule be a WP:SYNTH issue
 * coule be a WP:SYNTH issue

Data
Data retrieved from stats.wikimedia.org. Copy and paste table into a plaintext file to keep relative column numbering. Data used in the figure runs from January 2001 up to December 2012.

Data
Data found in this table from 2011, tab-separated version used in this script available at github.

Comments

 * I did wonder why this kind of analysis hadn't be done with the data, but maybe it was and it was seen to be not very interesting!