Making figures of your research can be gratifying in the "Gee, when you put it all together like that it's kinda impressive!" sort of way. It can also be tedious in the "Criminy this is tedious" sort of way.
What aspect of figure-making would I term the tediousest, if I were inspired to abuse my mother-tongue? The candidates:
* Gathering together all the data that despite your best organizational efforts is scattered across three binders, two lab notebooks, and seven computers, one of which has a dead hard drive and another of which you've forgotten your special login/password combo for.
* Finding figure-quality examples. You know the data add up.....so how come no individual example is ever "representative"? A mystery for the ages.
* Scale bars. Even when you have the Photoshop file with pictures of your calibration slide, there is always some sort of problem: you've changed the resolution of your picture, you've montaged several pictures together and they're no longer the same pixelage, your old camera died and the new one takes slightly different photos and no one has updated the Photoshop calibration file....
* Realizing that you analyzed your older data differently from your newer data and you have to go back to re-analyze months of work. Yum.
* Trying to make 5 equal-sized graphs fit into a multi-panel figure. "Hmm, 2x2 +1? Or, um, 3x2 - 1? Fuck it, I'm creating an unnecessary graph just to make it all fit right."
Of these many options, I surely loathe scale bars the most. Because even when you've gotten it all finished, you print out the graph triumphantly and then think...."That's odd, there's no way in nature that that object could be that size. Ohhell."
15 years ago
4 comments:
For me it's the gathering data part... I'm in data gathering from co-authors hell at the moment. Never before did I think it could actually take TWO WHOLE DAYS to make one excel sheet supplementary table. Never! But it did... and I still have more table to make
I'm having the same problem with a batch of EMs, many of which don't include scale bars. It goes something like: all right, so 3.78 pixels are a nm in this one, and this one has a scale bar, let me measure it... I need a drink...
this post made me laugh.
jenn,
2 days for a table is NOTHING. Try 2 months. Try a year and a half.
dr.j,
when I was a grad student I hated scale bars too. Now I do them as I go along, in fact I make mockup 'figures' from each data set so I can always make sure I have some 'representative' examples. It's MUCH easier that way. It still all gets changed around later, but it helps. Still have to go back sometimes for the Money Shot, but still better in the long run than waiting until the end to make figures.
I totally agree about layout. That has gotten somewhat easier but I don't have the skill to see how it's going to work out BEFORE I start resizing everything.
Somehow there is always one graph that's bigger than it needs to be, and a montage that's too small because the orientation is wrong and there's just no other way to make it all fit on the page. ARGH.
Ms PhD--making the scale bars as you go....that sounds like postdoc wisdom! And I have literally contemplated adding schematics or adding/removing graphs to avoid the giant white spaces in a figure.
We often use a confocal microscope whose (leica) software, I swear to FSM, makes it well-nigh impossible to save out a scale bar in any logical fashion. It's unreal. Who designs these things anyhow? Sounds like Jenny F's EM problems....
Post a Comment