![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
kareinaI am getting quite frustrated with this set of calculations. It should be routine by now, but I'm getting really weird results, and I just can't figure out why! There are two types of calculations I do with these samples. With the first I tell the program the whole-rock composition and it calculates, over a period of about 45 minutes, which combination of minerals will be stable at each temperature and pressure combination (within a range I give it before it starts), and what the compositions of those minerals will be. Once it does that I create graphs showing the changes in garnet composition for each temperature and pressure, and then highlight the lines which corresponds to the actual composition the core of my garnet. The point at which those lines intersect is, in theory, the temperature and pressure at which that garnet first grew. If all goes well with that step I obtain a printout of the compositions of all of the minerals for that specific temperature and pressure and the composition for the whole rock, double check to see that the latter actually matches the starting composition, and if so, I use those numbers to run a new set of calculations which starts with the temperature and pressure I input (which I found via step one) and then calculates the expected changes in the rock composition if the temperature and pressure changes along a specific path (which I input) and, as the garnet grows the center quits reacting with the other minerals and becomes frozen into a series of changing compositions. I then get it to tell me the compositions of the minerals at various points along the path, and compare the garnet composition changes with the changes in composition in the real garnet, to see if I've yet modelled the growth of this particular garnet.
Normally, as one might expect, the first point on the "path" calculations has the same conditions as the first step calculations (which they should, since they are use the same data for the same temperature and pressure). However, today, they are not the same at all. They are not only predicting very different compositions for the garnet, the path version is predicting completely different minerals! It thinks that there should be clinoamphibole present. Not only is clinoamphibole not present in this sample, it has never been predicted for any of my samples at any time. This is weird. The obvious answer is that I have a typo in the data I entered from which it is doing the calculations, but I simply can't find it.
So then I decided to double check the first set of calculations and try again to see if it gives the same results the second time. Nope. Not even close. Something weird is going on some place, and I have no idea why! And it is now 22:35 and I've not written *any* new words for my thesis today. This is not good. I'm now tired and frustrated, and torn between putting down the computer and going to bed, or finding some other aspect of the thesis to work on for a bit so that I can get some words written today.
But rather than leaving you with a rant, I will also report that I managed to accomplish a fair bit of non-uni stuff today. My kitchen ceiling is, once again, mould free (and it hasn't been all that long since the last time I had to clean that!). The wall behind the kitchen couch is, once again, mould free (though I must confess that it has been a bit longer since that task was last done). The floors have been swept & spot-cleaned or vacuumed (as appropriate), my clothes have been washed, and I made a bit more progress on converting that knit dress into a cardigan--the old turtle-neck has been separated into two halves and each half has been stitched down to the sleeves with the first pass of stitching. Next I need to cut off the original, weird, puffy sleeve endings and finish the seams at the join and sew shut the new portion of the sleeves. Then I can attach the puffy bits as pockets and add some sort of button closure system and I'll have a wearable garment!
Normally, as one might expect, the first point on the "path" calculations has the same conditions as the first step calculations (which they should, since they are use the same data for the same temperature and pressure). However, today, they are not the same at all. They are not only predicting very different compositions for the garnet, the path version is predicting completely different minerals! It thinks that there should be clinoamphibole present. Not only is clinoamphibole not present in this sample, it has never been predicted for any of my samples at any time. This is weird. The obvious answer is that I have a typo in the data I entered from which it is doing the calculations, but I simply can't find it.
So then I decided to double check the first set of calculations and try again to see if it gives the same results the second time. Nope. Not even close. Something weird is going on some place, and I have no idea why! And it is now 22:35 and I've not written *any* new words for my thesis today. This is not good. I'm now tired and frustrated, and torn between putting down the computer and going to bed, or finding some other aspect of the thesis to work on for a bit so that I can get some words written today.
But rather than leaving you with a rant, I will also report that I managed to accomplish a fair bit of non-uni stuff today. My kitchen ceiling is, once again, mould free (and it hasn't been all that long since the last time I had to clean that!). The wall behind the kitchen couch is, once again, mould free (though I must confess that it has been a bit longer since that task was last done). The floors have been swept & spot-cleaned or vacuumed (as appropriate), my clothes have been washed, and I made a bit more progress on converting that knit dress into a cardigan--the old turtle-neck has been separated into two halves and each half has been stitched down to the sleeves with the first pass of stitching. Next I need to cut off the original, weird, puffy sleeve endings and finish the seams at the join and sew shut the new portion of the sleeves. Then I can attach the puffy bits as pockets and add some sort of button closure system and I'll have a wearable garment!
