Sad but true, today is my last real day of work (tomorrow, David and I are going to DE to show off my movies... and maybe handle some other, "important" business). The movie that I'm making for Rich (Beta Cephei emission measure with twice as many magnetic field lines) is taking slightly longer than anticipated. I'd guessed 27 hours, but it's looking more like 29 or 30. Not sure if I'll come back in later tonight to redo all of the movies that I haven't done yet.

In other news, I continued work on my user manual/fake thesis. All of the images are in and in .eps form. I've also added citations where appropriate (boy did that articles page of mine come in handy). I've added one article to my articles page because I needed to cite it-- the article where Marc Gagne first publishes the periodic variability of &theta1 Ori C. Why? Because citations are cool.

10:21 pm: I came in tonight to finish up the movies that I didn't get to. Also, I finished my first round of editing of my write up. The newest version is once again on my code page.

12:04 am:Finished.

The movie of Beta Cephei with overlaid magnetic field lines finished up this morning, and I've posted it on my movies page. Sweet!

Also, I've continued work on my fake thesis/user manual. The link to it is on my code page.

Spelling will be the end of me...

I also had to recreate the image for ApJ. In fairly rediculous fashion, I managed to mislabel my axes. But that's fine, because they needed to be completed redone to match Figures 5.a and 5.b (mine will be 5.c, so they gotta match!).

First thing: I realized this morning that the movie I was making overnight (emission measure with overlaid magnetic field lines) was of the entire 18x18 region, so it was going to take too long and not be detailed enough. That run has been restarted now on the same grid as for the t1oc movie. Before I did that, though, I made the remainder of the movies, including line-of-sight velocity (for both; I redid the t1oc movie to get rid of the black stuff at the top) and the bcep emission measure grid.

I'll probably spend the rest of the day writing my users manual and making necessary changes that I find. I've now put the first version of the users manual up on the web. See the links to the left for the file. I've also put a link to the file on my code page.

Ugh. Dumb, dumb, dumb, dumb error. Turns out that every time I though I was using temperature, I was using one of the velocity vector components. Dumb. This means that I have to redo my EM vs T histogram as well as all of the movies that I've already made that involve temperature. Sigh...

Upon redoing one of those movies, I discovered that my line-of-sight velocity calculation can't handle being tilted. I'm probably just not handling the new third velocity vector component properly... UPDATE: Everything should be working fine now. As expected, I wasn't handling the third velocity vector component well (or, as some might argue, at all).

There are some new movies up though! Check 'em out! I redid the t1oc grid movie and put up temperature and grid movies for bcep. Also, I redid the EM vs temperature movie and am currently running the emission measure with overplotted magnetic field lines movie on the PC right next to me. Stay tuned...

One last thing... I restarted the manual/fake thesis. I'm using Lyx now, and trying to keep it a little more organized than the first effort. Hopefully that will go well.

Due to a power outage last night, Ruby was down for most of the day. When Chris finally got it back up and running, I managed to make a couple of new movies. First, I made an emission measure vs temperature histogram movies, as well as a line profile movie. I then moved on to the grid movies, but could only do the line-of-sight velocity themed grid since it's the only one that doesn't use the temperature contour. In the meanwhile, I finished up the image to send off to ApJ and put a version on my images page.

Something is seriously wrong with the temperature contours. For some reason, I get a lot of white space where I shouldn't. Now, white is the last color on that color table, so I figured maybe the contour levels didn't reach high enough temperatures. But after making the maximum temperature 10¹⁰⁰ K and testing to find that the maximum temperature was 10⁸ K, I decided that this was not the case. I then thought that maybe there weren't any data points out there, but when I overplotted the data locations I found this also to be false. So basically right now I have no idea why this is happening. It's strange, I mean the data's all in there and it looks consistent with the rest of the data, but the contouring routine just won't handle it right now.

My last act of the week was to begin making a movie of &theta¹ Ori C with equatorial magnetic field lines plotted. Last time I think it froze around 120 ks, so we'll see what happens when I let it run for 3 or 4 days.

I made some more movies, most notably of the beta cephei simulation. I've also added a line profile movie to both.

The biggest project of the day was making an emission measure histogram to put into ApJ. So far I'm having a difficult time making it into a postscript, but the jpg version is posted on my images page.

My magnetic field lines now successfully rotate. This applies to both drawing routines, both angular and equatorial. I've put up a couple of example images on my aptly-named images page. I've also ported these new methods to work on the gridded output in d3grid.pro as well, and it's been completely ported to d3t1oc.pro.

In addition, I've finally created a movie of &theta¹ Ori C with the magnetic field lines drawn overtop an emission measure contour. The movie can be downloaded here.

I'm now moving on to d3bcep.pro, a program to handle the output of the Beta Cephei run. So far things are going moderately well, except that for some reason I can't get the temperature contours to work. Also, I'm running a t1oc movie of the right half of the star, emission measure contour, with equatorial magnetic field lines overlaid. This should be done tomorrow (hopefully...).

In addition, I've created an all-new movies page!!! Now they're all collected and organized, with space for my soon-to-come beta cephei movies as well.

Two new magnetic field developments. I'll go in order of increasing coolness. First, I've enabled all of the magnetic field drawing routines to apply to the gridded outputs, only on the emission measure contour plot. This is drawn using the same /magfield keyword as before.

More interestingly, I implemented David's idea of starting the magnetic field lines on the equator to try to find isolated magnetic field lines that might result from the magnetic field of the plasma. So I chose 8 equally-spaced points on the magnetic equator and modified Marc's code to do basically the same as it was doing, only now starting at some point on the equator. The results are fairly interesting. As my images page shows, there are in fact instances of isolated field lines.

Of course, this new algorithm can be implemented everywhere that the /magfield algorithm can by calling /equatorial instead.

I spent most of the afternoon trying to make it so that the magnetic field lines would plot properly if a non-zero tilt angle is entered. While I was successful in transforming the magnetic field vectors to plot properly, I was highly unsuccessful in trying to pick which points to start drawing at and at determining the boundaries.

Today I spent all day trying to fix my magnetic field contours. First, I tried some IDL-defined functions, but they were only good for plotting vector fields and not field lines (in other words, they would plot a bunch of little arrows but not connect them). Naturally, then, I tried to come up with my own routine... bad idea. After an hour or so, I decided that it wasn't worth the headaches, so I busted out Marc's old code, which can do magnetic field contours successfully. After some tweaking, I got it to work in zeuscontour.pro, which of course only shows one half of the star. I ported that code to d3grid.pro and am currently successfully plotting the right half of the star. The left half, however, is still giving me some trouble. I need to really go through the bits of Marc's code that I'm using and really understand them so that I can properly map them to the other half of the star.

On extreme downside is that the magnetic field contouring takes on the order of 10 minutes per time snapshot. That's going to run to about 14 hours for the magnetic field contours, plus the hour or so to do all of the other processing, so once I do get this working, creating a movie is going to be an overnight affiar.

I've made some more movies. See the links to the left to download them (right-click and save the file before playing). One of the movies features my new emission measure-themed grid. Another features my new line-of-sight velocity-themed grid.

I also did a little test to make sure that my line-of-sight velocity calculations are correct. I took a spherically-symmetric distribution (I used t1oc at t=0, since this snapshot is just the initial, spherically-symmetric conditions) and found the line profile statistics at various tilt angles. The numbers that I found are summarized here. While they're not all exactly the same, they are extremely consistent and are always within 1 percent, which is about as good as I can expect considering that my volume calculation is good to about the same accuracy.

I think that all of my histograms are finally formatted the way I want them. I fixed both EM vs Time and VLOS vs EM histograms in both d3t1oc.pro and d3grid.pro for both being standalone output and being plotted as part of a grid. So that's done (I think). I still need to make a couple of different optional grids, particularly one that focuses on emission measure and one that focuses on line-of-sight velocity.

I wrote up a little bit of my code guide/manual/fake thesis. That was fun. Then I began working on the first of two new grid-type outputs. One will be EM themed, the other line-of-sight velocity themed.

Today is independence day for: Kiribati

I've realized that my code is getting a bit messy and a little inefficient, so I'm going to take the better part of the day to clean it up. The first thing that I did was to make starwhite.pro, a program which fills in the white circle over the star. This will eliminate about 200 or so lines of code from every program except zeuscontour.pro, which only calls it once, although I may implement it there anyway just for consistency.

I've now gone through d3grid.pro thoroughly and d3t1oc.pro somewhat thoroughly to make sure that everything is cleaned up and running moderately efficiently.

I'm now finally satisfied with the way d3t1oc.pro handles the EM vs. Time histograms, so I'll be moving on to the EM vs VLOS histograms.

So I finally fixed my LOS vel. calculations. Turns out that I was rotating twice and miscalculating the velocities to begin with; that means that every time I changed something, I fixed one problem but not another. So now it's fixed in both programs. I've also got my VLOS vs EM histograms formatted properly in both files, although I may not have fixed in D3grid.pro in the grid keyword case...

Today, Marc Gagne came and gave us the talk that he's going to be giving at a conference whose name I now forget (I'm writing this on the 11th since the CS network was down on the 8th at the end of the day). At lunch, I gave a little presentation about what I've been doing all summer, which I think went pretty well. I then spent the rest of the day trying to get my formatting correct on the EM vs Time histograms. I think I should just write my data to a save file so that I can do the histograms faster, then write my final formatting version into the code.

So David had a thought today, and usually that means trouble for me. He pointed out that I'm only working with 2 vector components for velocity (and for magnetic field, but we'll cross that bridge when we come to it). This means that my line-of-sight velocity calculation would be wrong for every slice excpet the one that I view: &phi = 0. This, in turn, will throw off my line profiles. Needless to say, a disaster.

In addition, our troubles with the magnetic field contours have continued. There's just no logical reason why the IDL contouring routine should do everything correctly except the magnetic field. So I have to assume that it's doing the contouring properly. At the same time, the only thing that I'm doing to it (to date, at least) is find the magnitude, which for the slice at &phi = 0 should be correct.

In the meantime, I made emission measure vs. time histograms using d3t1oc.pro. You can see the resulting images, as always, on my images page. I've plotted the same histogram with and without occultation simultaneously, as well as put a horizontal dashed line across the value of emission measure at time=0, or the theoretical value for no hot plasma.

Today I did a lot of cosmetics on my output images. First of all, I enabled the simultaneous plots of temperature, emission measure, line-of-sight velocity, and the line profile all in one image. This then translated into a new movie conataining all four. In addition, I changed my line profile so that it would have universal axes (i.e. unchanging between time snapshots). I also made all of the line-of-sight velocities in the histogram scaled to 2500 km/s - the theoretical terminal wind speed. Now, line heights and widths can be compared quantitatively visually, not just numerically, making my images much more intuitive.

Then David and I spent a good chunk of the day trying to figure out why my magnetic field contour lines don't look right. Still no solution there.

I made my first movies today. See the links to the left for the movies. These are direct results of my new program, d3t1oc.pro, which takes the full output file from the MHD simulations and, so far, makes movies of temperature, density, emission measure, and line-of-sight velocity. Soon I hope to add line profile movies, and make a 2x2 array movie with temperature, lilne-of-sight velocity, emission measure, and line profiles in separate panels that would be synchronized.


Today is Independence Day for: Venezuela