Happy Thursday. Great job with the move today, everyone! And the BBQ yesterday - go team!

As of now, I have a script that can be used to identify data point clusters in oceanographic data, save the cluster labels back to the database where the data is stored, and remove the labels. I can also do all of those things sequentially by stepping through the data and identifying clusters in each step. I hope that sort of makes sense...

Anyway, I've reached a good break point to start working on my final paper and presentation.

I've also been making some changes to my set-up so that my software will run more quickly on my local server. This is been a process but I think I might be in good shape now. Just waiting for a big database to load. Cross your fingers.

I think that is all. Work has been a bit slow this week because of everything going on.

Have a lovely weekend, see you all at the Open House.

Cheers,
Rachel

Hey guys!

Crazy week. Great job on the BBQ! Due to everything going on this week, I havn't had much time to work on VARS stuff, so I am still trying to get through pulse 66 but I should be finished with it by the end of this week. After that I will start on the cross correlation data analysis. Hopefully I will get everything done just in time for the presentations! Hope everyone is having a great week and enjoying the new housing.

• Danielle Fabian

Aloha e,
Hi everyone! Awesome BBQ yesterday. Many people complimented the food and how we all worked together nicely. I really cant believe we have two and half weeks left. I am finalizing my design things and trying to get everything finished up. I am still waiting for video from Clarissa in San Diego to finish my video projects. I cant wait to do the tour of NASA, how cool is that gunna be!!!! The move went well too I think. I dont have much to report, except I am on track but still need to work hard and pull it all together. Looking forward to seeing the results of everyone's projects but sad the summer is winding down.
Mahalo
Alia

Land sickness is no joke...

As if suffering from sea sickness wasn't enough... The Flyer cruise was AMAZING! I learned so much and got to participate on a bunch of different projects while at sea. What I didn't do a lot of was work on my project. I had grand aspirations of writing up my paper, but the meds I was on dilated my pupils, which made focusing on a computer screen a bit difficult. We did see a bunch of Procymbulia sp. and even some with some mucous feeding webs! There aren't many annotations of the webs because they usually ditch them when the ROV shows up. We were able to sneak up on a couple and take some frame grabs before the webs disappeared. They will help out with the line drawings! Now, back to writing

Hey everyone,

It was a fun week this week with the birthday cake break, quick COSMOS discussion and BBQ prepping/cooking.  As far as my project goes, I finished collecting all my salt water data.  This means I won't be using the Raman laser anymore (it's such a shame because I was finally starting to get the hang of working in the dark) .  Running the MATLAB baseline corrections and Gaussian fits is also complete for salt water, so now I'll just be sifting through some papers about temp/pressure/salinity and their effects on various chemical properties to get ready for the paper and presentation.  I'm also looking forward to the Open House this weekend with all the different activities and stoked to see the inflatable vampire squid.

 Cheers,

David

Hi everyone,

Crazy busy week, but a good one nonetheless. I finished the latest version of my write up which includes two different time periods in which I compare the three moorings I'm working on at different times of the year. I had some troubleshooting to do, and some plots from early on I had to fix to include more data, but it came out pretty well and the seasonality within Monterey Bay, and its interaction with upwelling, that was present was really interesting.

While I'm sure I'll keep coming back to these analyses I've been doing and refining them, I am now starting a new part of my project. I will be focusing on pCO2 in air at each mooring, particularly because from my analyses it was clear that during periods of relaxation or purely South winds pCO2 in air at Hopkins, and to a lesser extent M1 and OA2, increases. The idea is to see whether this has any net effect on the local scale air-sea fluxes, which is interesting and hasn't been studied before although it has been known for many years. Thankfully I still have 4 weeks to go, so I'll let you know how this new endeavor goes in the next update.

Have a great weekend,

Diego

Greetings!

 This week has been weird with everything going on and the end of the project looming. I started by sequencing the last of our PCR gel cut products, to make sure that I had a nice, clean transcript that I can clone in bacteria and express later. I designed some cloning primers on Monday, which is a lot more complicated than making amplification primers because you need restriction sites, extra bits to make sure the replication enzymes can bind properly, and it needs to start and stop exactly on your start and stop codons, so you're not cloning extra genetic material. That's why we needed such good sequence, so we could be sure we knew where our open reading frame for our potential protein was.

 The primers got here today, so I'm setting up another PCR with my new cloning primers, and will be digesting and purifying both my DNA and my cloning vector later today. The next steps are to transform my fragment into the vector, then into my E. coli, grow up the bacteria, select colonies, purify/sequence (to make sure we actually cloned the vector sequence), and then transform the vector into a different strain of E. coli that will hopefully express my protein!

If everything goes PERFECTLY, I'll have an expressed and confirmed luciferase protein from vampire squid a day or two before the presentation. Wish me luck!

Happy weekend everyone,

Chloe

This week I finished a data logging system that read from 2 sensor devices to get linear and angular acceleration along with compass heading from a magnetic field sensor. So what I have is a mobile system that can be mounted on something and records this data onto an SD card. The data recorded on the SD card can then be processed on a desktop, this is not the most ideal set up for the eventual goal of automating CPF drift correction but it is a solid basis for determining the feasibility of such a system. Now the hardware is working and the sensor drivers are written, I've started processing some output data in MATLAB to try and determine whether or not I'll be able to measure the types of accelerations we should see on the CPF. The likely answer is that we wont with the sensor being used right now, however the sensor was chosen specifically because there is another (much more expensive) sensor made by the same manufacturer and using the same programming interface that should be able to do it. The same software I wrote for the current sensor should run on the fancier one without any serious modification required.

The hardware programming was fun and challenging, I'm a little less excited about the data processing considering my pessimistic expectations of the performance of this sensor but time will tell.

-Pierre

Hello!

I have had a very successful week. There has been a bit of frustration with library dependencies for the last of the code but we were finally able to run the Inception v3 model and got some really great results. We got 100% accuracy on the blue whale true-positives from the network but only a 20% accuracy for the false positives. I am re-annotating all of the spectrograms and praying to the computer science gods that re-annotating misclassifications will the network happy. I've been trying to get started on the report and presentation but I feel like I can't start that until I get some results that we are happy with.

BBQ was awesome. You guys killed it. I'll be enjoying leftovers for a good minute.

Have a great weekend and don't forget the open house!

Daniel

Happy Thursday!

Great BBQ yesterday team! there were lots of happy well-fed MBARIans , yay! this week's been very busy, as this internship comes to an end. I spent the majority of my time obtaining data outputs and writing a code to be able to compare backscatter concentration to certain parameters. I've been able to independently work on my code this week, a great accomplishment..keeping my fingers crossed that everything goes well the rest of the week, I will begin analyzing and interpreting data next week, and begin writing my research paper.
As for the REU, we have an oral communication workshop Friday (tomorrow) and a mid-summer reflection writeup.

hope you all have a great weekend!

Ash

Hey Everyone!

Bit of crazy week with the cake break and the barbecue! I finally figured out how to work with R on my logistic regression and how to interpret it so that's exciting! And I've spent the rest of the week working in VARs and getting as much possible data before I have to analyze everything. The barbecue was absolutely insane but I had a ton of fun working with everyone! Thanks so much to David who went along with Danielle and I to buy everything (3 stores later and getting home at 9:00 pm but I had a blast!) It seems like it turned out great so I'm super happy about that . Spent the rest of today not in VARs at the Land Use Watch Panel which was once again a lot of fun. That's about it - I can't believe we only have two weeks left!

Have a great weekend everyone!

Victoria 

Oops, late confluence!

Had a rough week of some procrastination as the big deadline is coming up, but I'm back on track now! Some writing is done for ESP, I have edited the BTS news articles and sent them out to Benoit-Bird to review, although I haven't heard from her yet. I'm going to pop by her office if she doesn't reply to my emails by today just to make sure everything is okay.

Remember to send in a guess for the mystery picture! http://www.mbari.org/animal-plant-mineral-or-man-made-5/

Looking forward to open house!

Emma

This week processed two new videos from the video lab. This will give me more data to test and train my algorithms on. I also decided to go back through my annotations and add a few animals. My networks were getting confused with a some similar looking animals such as the Tjalfiella, Oneirophanta mutabilis complex, Fungiacyathus marenzelleri, Cystechinus loveni, and the Synallactidae gen. et sp. Indet. I'm hoping that annotating these animals will allow my network to better detect and classify the animals I really care about as I will remove these in post-processing. I am also able to make pretty results graphs that compare the number of training images to the average precision for each animal. Also grilled ~200 hamburgers.

Week 8:

Yesterday, I had to stay a bit later after work to collaborate with one of my adopted mentors, Danelle, and see what kind of hardware connections were needed for communication between SeeStar (Sleepy Pi) and Feather. Should it be TTL to RS232 to RS232 to TTL or straight from TTL to TTL between the boards? I ended up recreating my current set up for Danelle, which meant soldering header pins onto the Feather, wiring pins, and of course polishing and uploading the code.There was a lot of debugging via code and wirings, but we discovered that somewhere along the connection between SeeStar and Feather, the Feather fritz-ed, and was no longer sending the proper bytes, but instead garbage.

We tried out my current Feather set up with a simple test, which writes the character 'U' and it worked just fine. Next week we will finally test the protocol I wrote and see how SeeStar can receive and send information with the Feather. 

As for the software side, the Feather reads two basic opcodes, $1TSYYMMDDHHMMSS and $1TL. The first one, (Take Sample)  orders the Feather to read samples from all its sensors and log the information onto a file with the Year, Month, Day, Hour, Minute, and Seconds as its timestamp. This means that the time is relevant to the SeeStar even though the Feather has its own clock. The second (Take Log) means to send back the last information recorded onto the file back to SeeStar and can be used to check if reading and writing commands are functioning.

It was a lot more effort trying to integrate previous work with my own. Even though my present code works, it is messy and inefficient, whereas the code I wrote on my own was naturally much cleaner and coherent. Although the work I put into it does not reflect the quality of the product, I am glad that I went through the experience of learning someone else's protocols, understanding their flow of logic, and trying to meld together ideas to fit past accomplishments with currents needs. This type of collaboration is common for software engineers, so it was beneficial for me to reflect how well I do.

Working with other engineers has also given me firsthand experience on projects. I have noticed and have even heard other engineers admit how much their ideas differ, despite designing the same product. As we collaborate as a team, we do not essentially feed each other affirmation of our ideas, but rather clarify problems at hand and question which of our methods answer them the most. 

Another thing I have enjoyed here is seeing how much experienced engineers still go through similar troubleshooting difficulties as someone at my level. I often worry that as I progress academically and professionally that I should not be making or encountering "trivial" difficulties, such as establishing serial connections or incorporating libraries. It is nice to know that despite years of experience, people still make mistakes. 

Besides polishing up my code and testing the protocol with the SeeStar, there really is not much more work to include. I felt pressured to write a complete, clean protocol and share a final product for the SeeStar during the test yesterday, but Dannelle admitted that prototypes are often messy, which alleviated a lot of my stress. I can possibly add a check sum function, which adds the bytes of all the characters sent and compares it to the last byte for the purpose of data integrity. I think it is similar to parity bits, but in this case bytes are summed not just bits.

My priority for the rest of this week and next is the presentation and paper as I will be having a mock presentation next week in front of the SeeStar group!   

delete

Hi everyone!

What a busy week! Cookies were baked, tons of boxes packed, and air mattresses blown up!

I've had a fairly hectic week at the lab as well. When I came in on Monday, my mentor told me that we'd run out of sheath fluid for the flow cytometer over the weekend, so we didn't have any data on growth rates, size, or fluorescence from Saturday-Tuesday. We preserved samples to run when the sheath came in, but on Tuesday we had issues running them on the cytometer. We decided to save them until we were certain we didn't need to use the cytometer for an ongoing project. From Tuesday onwards, we've been working on wrapping up the growth-curve experiment. I've spent a lot of time running photochemistry tests, and today started working with my mentor to analyze all the data we've been collecting. We're continuing to run the CO2 and nitrate experiment in the chemostats.

I'm giving a practice talk to my lab at our lab meeting on Wednesday, so I'll spend most of the time before then working on getting that solid.

Hope everyone has a good weekend!