Den Wald vor lauter Bäumen nicht sehen …

A common German phrase reads: “not to see the wood for the trees”. I guess it is also know to non-German speakers? I just found this very trivial and easy Matlab command line, which makes my equivalent lines look very stupid and complicated:

a= [1; 4; 6; 8; 12]; k=9;

[min_difference, array_position] = min(abs(a – k))

These treasures make my day!

Fourier analysis for dummies

(Click on the image to be forwarded to the page.) Thanks for Alex for bringing this to my attention.

Subplotspace

If there is something like the most loved Matlab routine from the File Exchange side of Matlab then it is subplotspace. Subplot figures are nice, except for their huge spacings in between the subplots. This routine is very useful in reducing the empty spaces. Try it you will love it.

LaTex troves

You probably have experiences several of the moments where you have tried to find a solution for a simple/stupid problem in Matlab or LaTex. From now on I will try to post them here from time to time to share these troves with the other nerds out there. Enough baby stuff for now, back to the real life

Ever wondered how to insert a figure in a two or multi column paper, without being creative? Thanks to this guy who showed me how to do it:

\begin{figure*}
\end{figure*}

PS: Almost everything one has to know about figures and LaTex can be found in this German manual: ***** (PDF file)

Resume of wavread(‘baby’)

Slightly more than 3 month have past and our son has not shown any development in the frequency spectra of his voice since my last check. However, despite the spectra and time series, the sound is much more bearable than the last recorded sound track:

Waste of time

More than a month has past since I started worrying about the huge file sizes I got with MATLAB when saving figures with huge datasets (is that even an English word?) as EPS files. Luckily I had the time during my parental leaves to search and seek for a solution. The solution for now is

a) instead of pcolor use contourf, which amazingly reduced the file size a lot.

b) seperate the fancy figure into a bitmap and ps file, where the first contains the fancy image and the latter the text, axis, labels, etc. and then combine both to yield a new EPS file, which can be edited by LaTex afterwards and is as small as a normal high quality JPEG figure. Luckily someone has written a short MATLAB routine for it.

c) avoid fancy 3D surfaces plots, they look nice, but are not suitable for papers or reports. They are rather helpful for PowerPoint presentations and then simple bitmap figures are better than EPS images.

Now, after several parental leaves and many hours of neglecting my son for stupid figures, it is time to rule of (Schlußstrich ziehen?) and go over to the real problems, namely interpreting the figures.