Projects

Feel free to check out my GitHub page where I have a few projects uploaded by clicking on the cat icon.

Moving to Continuous Deployment - Lessons Learned

HOVER’s journey to Continuous Deployment has been one of experimentation, failing, learning, and success. This talk will describe the transformation that was necessary for the company, engineering department, and humans that work there to be more successful. I will discuss the process of moving from slow two-week deploys that had a high probability of being risky to continuous multi-day deploys that were more stable. I will walk you through the change in the engineering culture that involved testing first, new project structure, learning new skills like TypeScript, Docker, and building a more robust Feature Flagging practice.

This journey has been filled with lots of unknowns, and I’d like to share how we got unexpected results both good/bad, the challenges to get buy-in from leadership/product, the huge shift in culture, the opportunity to incorporate new technology, and what we learned along the way.

Women In Tech Summit
Speaker: Erin Frolli
Abstract for Moving to Continuous Deployment- Lessons Learned
Slide Deck

OptiREF

Is an ongoing open source project designed to create a single platform where researchers working on real-time quantitative PCR (qPCR) projects can choose their optimal reference genes. Currently, there are three established methods for choosing reference genes using different algorithms; NormFinder, geNorm and Best Keeper. However, the original software packages utilizing these equations are not in a standard format that would allow for simple comparison across multiple methods. More recently, R scripts and library packages have been created that utilize one or even two of the methods, but not all of them together. There is a free webpage that allows you to upload your data and it will run all of the most common methods, however this resource is quite limited in its abilities and does include the effects of the efficiency values of your qPCR. As such, the results that are displayed by the program are not optimized for a specific assay. For OptiREF, we are creating an open source R library package that will allow anyone to choose or run multiple methods and easily compare the results. We will eventually implement this into an R Shiny app that will allow both programmers and non-programmers alike to use one platform to run the same analysis on their data with minimal effort by the end user. The working code can be found on GitHub Here.

PickleJS

"Cucumber with Brine" PickleJS is an addon for the Cucumber plugin for Cypress allowing automatic story writing. This open-source project was created to allow PMs, QA Engineers, and FE Engineers the ability to work together to create seamless testing plans.

The working code can be found on GitHub.
The library can be downloaded from NPM.
The Documentation for PickleJS can be found here.

Fine Scale Magnetic Field Maps

For the Weddell Seal project, we are testing the hypothesis that Weddell seals have a magnetic sense. In order to test our hypothesis we had to have multiple measurements of the magnetic field, because we were unsure at what level of detail the seals could detect. We had access to various past aerial magnetic surveys through the NOAA Marine Trackline Geophysical Data, however to ensure that we had the most update and accurate magnetic field information we mapped the magnetic field ourselves. Using a Geometrics G-858 portable magnetometer and a G-857 base station magnetometer we map the fine magnetic detail of three areas of the McMurdo Sound, Antarctica. We conducted our surveys during the October -November months of 2014 - 2016. The data and metadata for each location can be downloaded from the NOAA Marine Trackline Geophysical Data Viewer:

• B-017 Site 1

• B-017 Site 2

• B-017 Site 3

OptoGraphs

Is an ongoing project designed to create two simple assays for testing the optomotor and optokinetic response of larval fishes. Currently, known methods for creating graphics to test visual response of larval fish utilizes code for paid licences programs like MatLab. I was tasked with the job of creating a simplified and free alternative for the students in my lab to use for their undergraduate and graduate projects. OptoGraphs is a series of R scripts utilizing two methods for creating .gif graphics using two open source resources (ImageJ and ImageMagick). The working code can be found on GitHub Here.

Women In Marine Science

A custom website was created using R Shiny for an event sponsored by the Marine Science Institute at UT Austin to inspire young girls to be women in marine science. The website is designed to provide a visual of how data was used in the Weddell Seal project, including 3D dive data. The website is interactive and allows the user to select a dive and the data plots for the seal’s depth or speed. This selection will then change to show the data behind that specific dive. The site also has a 3D graphing page which demonstrates a single 3D graph of a dive that displays where the animal ate various prey items. Once the user selected a prey item, information about that item appears as well as a video begins to tell the user more about that species. Feel free to check out the webpage Here the code for this page can be found on GitHub Here.

JellyTrack

Jellies are one of the primary prey for the critically endangered Leatherback sea turtle. Adult Leatherback turtles travel from nesting beaches in Indonesia across the Pacific to the surface waters off central California. The adults will spend 2 to 3 years roaming from California to Hawaii feeding on jellies before returning to nesting beaches to mate and reproduce. For my undergraduate honors capstone, I created detection algorithm to estimate jelly populations in the surface waters off central California using aerial photography. This included developing a simple camera system to take the aerial photographs during aerial transects, creating and refining a detection algorithm, and testing the results against aerial observer observations. This method improved the estimate of the surface water jelly populations. To further this research, my graduate professional paper, incorporated this new methodology to improve the estimate of the jelly populations during a two-year period. My professional presentation is in the video on the right, you may also view it Here. My papers and supplemental data can be found on GitHub Here.