Montag, 4. Mai 2009

04.05.09: A day of rediscovery

  • Refactor flowDeconvolutor package, adding try-catch parts
  • Read and digest time curve analysis
Recently I am deceived by temporary and trivial successes too much, and therefore become almost too proud and seem to know everything. It is absolute absurd and with a day of really hard working it comes soon to the point that there is much too learn. I gonna tell me everyday: never give up learning, never close your mind.

Today I refactored the flowDeconvolutor package by hiding exception handling codes from user interface. The resulting codes become much more clear and compact. The next step is to consider the parameter dist and implement other algorithms. During free time I gonna write the documentation so that it can be submitted to Bioconductor project officially, hopefully next week.

The long-waiting RTCA story has now a new step: I made the figure of WST assay readouts of validated siRNA knockdowns, which show interesting dynamic patterns in the RTCA assay. During the plotting, I learned a new function in R called arrows and its couterpart in package named panel.arrows. The cool thing is the function helps to draw error bars instead of three segments. (It is really cool since I have never thought of the idea to treat error bars as a special kind of array).

By the way, it is also new to me there is a demo function in R (since in most cases I am used to vignette function in Bioconductor). Try:


, to understand how to draw error bars elegantly with lattice.

The figure at the head is given by the following code:

x <- stats::runif(12)
y <- stats::rnorm(12)
sd <- rnorm(12, 0, 0.08)
i <- order(x,y)
x <- x[i]
y <- y[i]
plot(x,y, main="arrows as error bars",type="l", col="orange", lwd=3)
arrows(x, y-sd, x, y+sd, length=0.1, angle=90, code=3)