I have succeeded in rescoring the key in it's new state, which is more flexible with regards to body form. This took somewhat longer than expected but was completed by Tuesday afternoon. Once the key was rescored I had Steve take a look at it, and his more discerning eye pointed out a few areas that could be improved to make the key more accurate and easy to use. I am presently implementing the changes he suggested, which are for the most part, minor.
Once these changes are done then I believe the key will be in a relatively complete state. The hope is then to send this prototype to noted ctenophore taxonomist, Dr Richard Harbison to get a second opinion on its reliability and structure. While awaiting his comments I will begin my presentation and report for submission in a fortnight's time.
was lucky enough to take a trip up to UCSC this week and was truly enamored by the campus, it seems like a wonderful place to learn. Also slugs and deer.

I have completed most of my laboratory work. Today I dissected the last batch of specimens. Hope to complete these tomorrow. Also I have started to work in images with imageJ. I am trying to find a method for counting age bands. As there is lot of seasonal bands in my specimens, it's confusing some times. Anyway hope I can figure out everything within next week. This week is really busy.

So this week I finished most of my analysis, wrote bits of my paper, and am working up a title.

Right now I have gotten as far as "Squishy Things in the Deep" but somehow I don't see that going over very well.

Next week will be more finishing up and doing some fun things with a microscope if we have time.

This week essentially saw the end of my project unless things dramatically change within the next week. There has been quite a bit of ups and downs over the past few weeks and it appears the necessity and goal of having the Wave Glider ready to go for deployment by September has faded. Currently the OA1 mooring is producing some funky data, so it has become the priority of the BOG to prepare another OA mooring for deployment by the beginning of September. This affects my project because some of the key elements that had to be designed or manufactured by next week will not be taking place. I was hoping to run the Wave Glider OA electronics as part of my presentation, but it looks like that won't be possible. The code for the OASIS controller (which essentially runs and communicates all of MBARIs electronics) was written by Bob Herlien and it seems to be acting a little funky right now. Furthermore, we have added a CO2 standard tank to the Wave Glider drybox. This requires an additional valve, which means there needs to be code written for the OASIS board to communicate with the extra valve. This needs to be done by Bob, however he might not be able to do that as soon as hoped, especially if it is no longer a priority. What seemed to be a pretty straight forward project doesn't seem like it will be completed, however I certainly learned quite a bit about the OA sensor system and what it takes to put together such a system.

This week I fiddled with the Wave Glider communication through the Iridium burst communications. This was established by creating an account through Liquid Robotics. The GPS also worked, as did intermittent recordings of location, speed, temperature and battery life. The system did act a little funky, however it may have occured because we don't have all of the needed parts. This week I also worked on my paper, which is nearly complete and I also submitted an abstract to the AGU meeting taking place in San Francisco in early December. I hope to present a poster about the MBARI integration of OA sensors into a Wave Glider titled, "The Wave Glider: A New Autonomous Surface Vehicle to Augment MBARI's Growing Fleet of Ocean Observing Systems." Next week I will finish my report and presentation and will hopefully also help build the new OA mooring. And perhaps, out of some miracle, I may be able to get the Sensor package working! 

This week started with an interesting road trip along the San Andreas Fault, including delicious lunch and some high quality gawking at crooked houses and fences.Over the last few days I have been working on an outline and abstract for my paper. If we can complete an abstract today I may be presenting my internship work at AGU in December. Such are deadlines! I finally have gotten my satellite temperature converted to pH codes to run, and next week I will be thinking about how to apply that visual tool.

In a poetic way the week will end tomorrow in Tomales Bay (the end of the San Andreas fault), where I am going to explore a shellfish farm. I will be looking for ways to tie the theoretical research on ocean acidification to the practical application of shellfish farming. 

This week I mostly worked on analyses and making figures for the paper and presentation. I finalized all the data for the mitochondrial loci, and so I was able to run AMOVAs for the complete data sets, and make tables from the results. I was also able to make the final haplotype networks for these two loci, and make them pretty in Adobe Illustrator. I also used Illustrator to make pie charts of the haplotypes found at various latitudes and place them on a map, and it was very interesting to see the gradient. I am currently running another analysis that Shannon thought would suit my data, and some preliminary results looked very good. It is on a program called IMa2, and it looks for patterns of migration between different populations, and time since divergence.

Today I volunteered for the Young Women in Science program, and it was really fun to see young girls getting excited about learning and biology. I also started one more attempt at getting data for a nuclear locus. We got results previously with a primer set for an intron in a heat shock protein (HSP), but the sequences were not quite clean enough, so we are now running them in the thermal cycler on a longer program in hopes that this results in a cleaner product. Next week I will find out if that works, and if it did, I will try to knock out all my samples as fast as I can, and work on the paper and presentation when not in the lab.

This week I finished processing the Raman data from my ternary gas standards, which serve as the last check against the binary standards I made up earlier in the summer, and are crucuial in determining whether there were any unforseen reactions in the impure gas. It appears that these are minimal, thankfully, and everything appears to have checked out well. In addition to that, I have been packing up pallets with Peter Walz and preparing for my talk in a few hours. I gave a trial run to my lab yesterday, which I was incredibly nervous for, but it went fairly smoothly and I have done my best to integrate their comments.

 Next week I will be on the Western Flyer off the coast of Oregon/Washington/BC looking at real life hydrates! I am incredibly excited, but it also means today is my last day at MBARI. I will miss it immensely, and thank you all for making this summer so wonderful! I'll see everyone at noon!

This week I was finally able to make the taxonomic tree fully expandable, collapsible, and searchable (with pressing enter to search and highlighting the current result). I was also able to make it so any search box can either do the search based on the page it's on or any other type of search if a different nav bar link is clicked. The data use policy page is working as well. Brian modified my representative images algorithm to make it faster, which is good for the higher level concept pages. I downloaded screen capture software and began experimenting with recording the DSG's many functions (which helped me find many bugs...).

In this next week, I will finish a couple minor things in the browse tree, including only re-expanding a node to only the first child and adding an "expand all button option. I will also add columns to the comparison table that we mentioned last week. I have yet to hear back from the video lab whether or not to implement a menu to customize the columns, or just have all the columns show up as the default. I will also work on my paper and the video of the functionality of the DSG in case I cannot manipulate it live on the Pacific forum computer. 

The geology field trip on Monday was great, I'm really starting to appreciate what my geologist housemates have been going on about all these years (usually I go to their talks out politeness, smile and nod the whole way through).

This week I have been using the Benthic Terrain Modeller to zone each of the canyons into various topographic groupings (flat, slope etc). This was a bit trickier than expected because all the the tools are design to work on relatively level terrain, the inherent slope of canyons confound it a bit. However, after a bit of tweaking with zonal definitions this was sorted.

Following this, I am identifying and extracting p/a and environmental data from each of my subsamples on the transects, this takes a fair bit of time but once it is done it's just a case of plonking my data into PRIMER for analysis. The PERMANOVA add-on for PRIMER should hopefully arrive today but if it doesn't come I have managed to work out how to run the test using R, slightly more complicated but needs must.

In addition to what most of the interns have expressed in their views on the Geology field trip, I must say I was slightly taken aback by my own genuine interest on a topic I had limited knowledge. I thought the way Gary presented his knowledge and expertise on the local geological area was very impressive!

This week I have been learning how to construct a confusion matrix to put all my data in a more manageable and easier-to-read format. I have results, but not happy with them so I have re-processed my entire data-set with a new training library (a more 'spruced' up version of the one I had) using the batch classifier. I ran the same training library with my random sample set and so far the values look a little more promising.

I have climbed a huge learning curve for me in terms of understanding the somewhat complex concept behind testing accuracies of classification systems in artificial intelligence. I plan to plot ROC curves which I have never done before, so that will also be the next step in terms of learning something brand new!An ROC curve basically tells you how accurate a particular classification system is, mainly derived from values from the confusion matrix. It will hopefully evaluate the standpoint at which AVEDac is in terms of its development.

On the side, I have been tweaking my Intro and Methods section which I think is complete. I have also been drawing up an outline for my presentation.

Next week I plan to finalize extracting my results, write up my Results, Discussion and Conclusion and work on my presentation. I also plan to end this madness I am having with the French regarding grad school which is also looming!It appears I am lost in translation.

This week, have not too much done yet, Collect some laboratory data but the MATLAB program was running slowly, I am still working on the programming stuff and try to do more Acoustic signature analysis in the end of week.
This week, I also worked on the paper a little bit and find out how to organize the paper, started to write.

Next week, try to complete the tank test and get acoustic data. Make good MATLAB program to deal with acoustics analysis. Complete the paper & presentations by next Friday, makes everything ready for the final week.
Not sure how far my project will be got, but I will keep working on that till the end of intern work. Good luck for me.

This week saw the collation of results for my paper. It is on track to have a draft finished early next week, which means that hopefully it will be ready for submission by the end of the month.

 I have also started looking at approaches for broadening the problem beyond just "front detection" to "water mass discrimination." Some early models using a simple K-means clustering algorithm have shown some promise. The primary problem with this approach is knowing a priori how many different water masses might exist in a given data set. So next week's problem will be figuring out a way to determine how to train the clustering system on enough different data sets to generate a reasonable set of centroids that can then be applied across all data sets. The advantage to such an approach is that it can be applied to any sort of ocean feature one might be interested in. It also can allow scientists to specify the parameters of what they consider an "interesting" water mass, and then the system can just apply the appropriate cluster to determine what to look for.

This week I spent a large chunk of my time researching the internal anatomy of cnidarians as well as working out the necessary steps and equipment needed for preparation of soft-bodied organisms for the SEM. I conducted some gut contents analysis on Atolla preserved immediately after capture aboard the Western Flyer, which yielded little results but the few that were observed proved interesting. Toward the end of the week I was informed I would be able to get back more histological slides of Atolla tentacles, so I prepped some more samples for the histology lab and dropped those off at CHOMP. At this rate the next two weeks are going to be jam packed with things to do in order to get ready for the final presentation and paper.
Hasta luego amigos!