Ok, that strange APNG error? Make sure the source file looks like this:
1.png; 1000; none; source
2.png; 1500; none; source
The final “source” appears to hold some magic ….
July 29, 2009
July 16, 2009
Thu: more R !! …
What are error envelopes: http://www.sr.bham.ac.uk/~ajrs/R/r-gallery.html?
#--Plot error envelopes for mean profiles, using "conf.reg" function # from "errorbar.R" script sourced above: conf.reg(CC,xtype="r.r500",ytype="sckT",fill_col="grey80") conf.reg(nCC,xtype="r.r500",ytype="sckT",fill_col="grey30")
Thu: R code update …
Cracked the generate a polygon — now just need to work out how to get it behind the main plot …
n <- length(hi.x)
for (i in 1:n) {
x1 <- c(lo.x[i],hi.x[i], lo.x[i+1],hi.x[i+1] )
y1 <- c(hi.y[i], lo.y[i],lo.y[i+1], hi.y[i+1])
polygon(x1 ,y1, col = “grey”, border = NA)
}
Thu: graphing pain …
Ok, so banding has something to do with polygon — now just need to sort out the vectors of the shading:
n <- length(lo.x)
for (i in 1:n) {
x1<- c(hi.x[i], lo.x[i], hi.x[i+1], lo.x[i+1])
y1<- c(hi.y[i], lo.y[i], hi.y[i+1], lo.y[i+1])
polygon(x1,y1, col = “red”)
}
July 14, 2009
Tue: R command of the day and further stats advice …
lowess(d.test$result)$y and lines(lowess(d.test$result), col = 2) are my new best friends …
Plan forwards:
- work out the loess for all results
- work out the average for all results
- show the variability of the results around the charted average
July 6, 2009
Monday: pratting around with graphs in R …
… and I am slowly realising where my life will be going over the next week …
It is evidently quite time consuming to work through the graphs, calculating the relevant pieces from each one. Sigh. Will crunch through them though, hopefully learning more en route !!
July 5, 2009
Sun: making head nor sense of charts …
Some functions to play with:
- pretty, which can be used to come up with axes quickly, e.g. pretty(0:-400000, n = 15)
- the minor.tick feature of the Hmisc library, for minor axis ticks
- how to install a R library http://bioinfo.cnio.es/faqs/5-faq-computation/8-how-i-install-a-r-package.html (install.packages(“clv”,dependencies = TRUE))
- http://www.statmethods.net/advgraphs/axes.html for general graphing goodness
July 2, 2009
Will notes 24-Jun F2F @ Rdg …
- Google “logistic maps” (Wikipedia reasonable) and “logistic map flocking”, the latter in relation to PSO, and cobweb diagrams
- look at Lypaunov sequences (?), will models converge
- Look a the Simon’s Computer stuff link for freeway models
- WB leaves –> NZ eof Aug.
