My project is officially done! At this point I'm writing out the report, presentation, and poster (for Hartnell). There's so much I have to do!

Also, NASA was amazing and I had a great time!

My project is nearing completion. I am currently working on quantifying values of interest for the swimming cycles of the bathochordaeus sp. I am hoping to be done with fitting a traveling sine wave to the tail motion tomorrow. Spent Monday on the R/V Rachel Carson doing deep PIV... it was amazing.

As week 8 comes to end so does my project. I am still having some problems with comparing my values with he values the code gave, so I am trying to find a threshold that works. Once I do that I can compare the orientations and then plot them all. I am looking to get this done by tomorrow, so I can focus on my presentation and report. Besides that, I went on the Rachel Carson which was awesome per usual and NASA was very interesting!

My code has been touched up, with the help of Mike, to shave off lines of code by taking those lines and making modules to make the code a little smaller and easier to understand. At this point the code can run a loop that will read all of the profiles for all of the floats that fit the criteria and from there I can insert the data calculations in that loop so it all can be done during that run. Aw yis.

Week 8:

Every week is so busy toward the end 0.0 Monday morning I took more SEM photos down at MLML, and my dad visited Monday and Tuesday! It's nice to show off how cool MBARI is and what I'm actually doing out here in CA. Tuesday, NASA was amazing, and I will definitely apply there next summer

As for my project, I'm still working on the projective reconstruction code. I'll have to start alternating working on it and filling out other subprojects since it isn't really imperative that I get the code fully working (it'd just be really cool if I did). Kakani is going to take a look at it tomorrow, and I think a set of fresh eyes will help a lot. In between, I'll work on taking and synthesizing ImageJ data of my beloved triangles, and mess around a bit more with genetic trees to see if there's anything more there to contribute to the silica conversation. Always so much to be done!

Looking forward to a friend visiting and the soccer tournament this weekend

Today, I did the layout for the SeeStar project and completed it because I was done with my project. I have completed the rough draft for my final paper and my mentor is reading it now. I have worked on the presentation and did final testing of the electronic load. I was able to solve the oscillation issue with a theory proven by my professor. I will be taking more test data if needed and modify the paper tomorrow.

This week has been awesome! I was sad to miss NASA on Tuesday, but that day I was able to finalize a calibration curve for my methane/ethane peak height ratios with an R squared of 0.99996 and error in the slope less than 1%, which is a threshold that I had been aiming for. Since Tuesday, I created a standard gas mixture to look at methane/CO2 and methane/N2 peak height ratios and did my final laser Raman analysis on the mixture this morning. I have the gas content of all my samples fully analyzed and have even analyzed some additional spectra from in-situ experiments as a comparison to the spectra that I took. Woohoo!

Week 8:

On Monday I worked with Mike McCann to discuss the ROVCTD data. The project seems to be rather large, however, so I will probably work on it on weekends and after hours. On Tuesday, Wednesday, and Thursday, I converted by data to micromol/kg (using the correct potential density values!) and have been finally removing outliers from my analysis dataset - this has taken much longer than I expected. I remove data if I have a good reason, and at times the problem is not obvious, for example the salinity or temperature casts have been off, thus giving an anomalous oxygen value in micromol/kg and especially when looking at density surfaces. Someone told me the BOG data was already clean ... thus my procrastination here and high levels of stress because I waited until week 8 to do this!! Tomorrow I will be playing catch-up again remaking plots with oxygen displayed in the correct units, making clean figures, cleaning up my dataset, and so on. I still think I can do some additional analysis - that is my goal, got to turn it up a gear though.

Nasa very cool, and an interesting seminar this week

I'm wrapping up my project, trying to make visualizations for the talk, which has been met with mixed success.

Still need to work on a summation of the physical environment but i'm not sure if I will include that in my symposium talk (i suppose it depends how much I get done).

Almost forgot about confluence this week!

My project has been surprisingly fun to work on. Beluga Whale!!! Bowhead Whale!? Narwhal Whale!? Whales!!!!

I'm at the point now to where I really need to pinpoint a species for determining cause for the depressions that I am seeing in my ROV video footage. After that I'm free to focus on completing my paper. More of a side problem I'm having right now is try to find locations for my depressions. If I can get lat-long data for all the depressions I can put them on some really cool ArcGIS maps. This id not something that is necessary but it can help to bolster my results and would serve as a nice visual during my presentation.

I worked a lot with QIIME this week and counted a couple more samples. We celebrated Alice's birthday this tuesday.

I got lots of work done this week, still tons more to do! Had some amazing meetings with people outside of MBARI- a professor from USCS and the seminar speaker, Caitlin. I am constantly amazed by all of the people I have been meeting and it has been really fun developing additional projects outside of the work I am doing for the internship.

This week has flown by. I have been organizing my data into what will hopefully be a presentation. My sensor will be deployed again tomorrow and it will be collecting data for the following ten days. Meaning that it is very likely that I will not be able to include all of my data in the presentation, and the paper will likely be a more complete look into the summer. I am still hopeful that I will be able to get my second sensor functioning before I leave.

Week 8
This week was all bout augmenting new data using the current dataset.The reason we need to augment data is because sometimes an overfitting problem occurs with the classification process i.e. the classifier might work really well with the current test and training data but fails if we bring in new acoustic data( this occurs due to small datasets) . The way I augment new data using the same dataset is to superimpose noise randomly from the hydrophone recordings I already have from different locations . As the recordings are from different locations in the bay , we get different noise , thus augmented data is more diverse. I implemented data augmentation with my current system and guess what , the classifier performance falls on an average by 0.1-0.2 percent which according to me is really good. The Random Forest Classifier works better than svm after adding the noisy data so I will use that for my classification purpose

Almost there! This week I made (with a lot of Brian's here) awesome graphics on Matlab that show every pixel in the demisions of the photo the ROV took in respect to the Canadian grid. So when looking at the graphic it shows that pixels at the top of the photo are small and more distance is shown compared to the bottom of the image where pixels are bigger and less distance is shown. These will be great graphics for the presentation and the paper we are hoping to publish! Speaking of presentation and paper, I have almost completed my presentation I just need to add a bit about the algorithm and I have also been making a lot of progress on my paper. The last step of my project is to tie in the data I collected in the beginning of the summer into the code, I will be meeting with my mentors today about this and hopefully I can wrap up the project this weekend/next week.

Could I have 10 weeks?. I would like to continue working on my project. Last week I thought my project was ready, but I needed to make some adjustments to my data. Nevertheless, this week I've been writing my paper and organizing my presentation. On the other hand, NASA was great!!

Super late update, but here it is. So last week we had an awesome tour of NASA Ames. I also spent a significant amount of time working on corrections to my scale chock design as well as scouting out the UHMW I will need to use in the machine shop. I also measured the kerf that is caused by the water jet during cutting to see if that would have a significant impact on the rightness and accuracy of the design. Other than that I gave a lab tour to the REU students Friday and showed off a few of MBARI's toys. Coming too close to the end, pretty nerve racking,