Google Summer of Code 2022 has recently wrapped up with 22 of our open source projects mentoring students.
Here are the projects who participated:
ArviZ, CB-Geo MPM, CuPy, CVXPY, Data Retriever, FEniCS, FluxML, GeoPandas, Gridap, LFortran, NetworkX, Optuna, pvlib, PyBaMM, PyMC, PySAL, PyTorch-Ignite, QuTiP, SciML, signac, Taskflow and Zarr.
You can also review a list of the ideas proposed by these projects for this session.
Keep reading as we share the personal takeaways and reflections of this year’s student cohort.
“I loved working with production-level codebases, iterating through design decisions, and reaching a logical conclusion with my mentors. The feeling of contributing to libraries others use for their work is incredibly satisfying.”
-Abhirath Anand, FluxML, Metalhead.jl development
“It was a wholesome experience that provided experience in various aspects of software development and professional work like version control, documentation, test-driven development and technical communication.”
-Abhishek Bhatt, SciML, Optimizing SciML Packages with SimpleChains.jl
“Gained experience in symbolic computation.”
-Bowen Zhu, SciML, Model Order Reduction
“It was an amazing experience where I learned a lot about how a lot of things work, as well as potential problems in the component packages during implementation.”
Chetan Vardhan, SciML, QuantumNLDiffEq.jl and Makie Recipes for SciML
“I learned the value of planning and other road mapping skills along with getting to explore various solutions all while enjoying the project I was contributing to.”
-Danh Phan, PyMC, Multi-output Gaussian Processes in PyMC
“I enjoyed the sneak-peak into the important things when being part of a varied group of people. My takeaway is that I want to align my future job to be able to continue contributing.”
-Greg Maya, PySAL, Street Network Simplification
“My biggest takeaway during the GSoC time was getting in touch with one of the most incredible teams all over the world, working & learning from them and improving my programming and debugging skills. I’m honored that my work brought an impact on the community!”
-Khushi Agrawal, CuPy, CuPy coverage of NumPy/SciPy functions
“I absolutely loved contributing to PyMC for my GSoC. Contributing to large projects can be daunting, but communicating with mentors and informing them about technical difficulties and also problems with timelines (work/school) will make contributing a much more pleasant experience.”
Kunal Ghosh, PyMC, PyMC — Fast Exact Gaussian Processes
“Outside of the learning objectives and the familiarity that I acquired with the codebase that I was working on, the biggest takeaways are 1) identifying a more tangible direction for my future career following my Ph.D. studies and 2) discovering the value of learning how to improve my workflow. I must thank both of my mentors, Brandon and Ricardo.”
-Larry Dong, PyMC, A PyMC Dirichlet Process Submodule via AePPL Enhancements
“Being a FEniCS contributor for GSoC 2022 has been the best professional experience of my life, because I learned a lot from my mentors, I worked side by side with skilled developers, and because giving tangible contribution to a great community as the FEniCS one is an unmatched feeling. Thanks NumFOCUS for this great opportunity!”
-Michele Castriotta, FEniCS, Expanding FEniCSx electromagnetic demos
“GSoC helped me learn new tech stacks, developing optimized solutions, and reaching out to other developers for help. Developing skills to be a good software developer and being active in the open-source community is really important as it helps one to be connected with the developers community.”
-Nagesh Bansal, Data Retriever, Data Retriever: Data retrieval using NeonVegWrangler
“Collaboration with people without any physical bounds is a great way to get different ideologies to be brought about.”
-Prayas Jain, Data Retriever, High-performance parallel computing for model fitting and prediction
“GSoC was truly a unique experience. I was able to learn new technologies, and overcome most of my introverted fears of communicating with others. I gained valuable life experience during my time contributing to PyMC. GSoC is an excellent opportunity for students to work on a project that is truly important while also exponentially improving their coding skills and ability to work in a large team.”
-Purna Chandra Mansingh, PyMC, Increase Support for Batched Multivariate Distributions
“It was a great learning experience to absorb more from the organization I was working with along with the programming aspect, too. It also encouraged me to engage more with the open source community.”
-Shashank Kirtania, PyMC, Create a Model class for easier deployment of PyMC models
“GSoC is the best way to get started with open source contributions. I learned a lot about how to write good and effective python codes, In addition to that I learned how to work effectively with Git and GitHub.”
-Shivank Chaudhary, Zarr, Registry for Zarr Codecs
“Discussing design philosophy and code organization with my mentors was illuminating for me. I’ve used open source code for years, but prior to this past summer, I had not spent time thinking about its construction and development. Gaining insights into how the Python Spatial Analysis Library and its modules were structured has made me introspect about the usage of code as an expression of thought.”
-Tyler Hoffman, PySAL, Interfaces for consistent statistical analysis in the Python Spatial Analysis Library (PySAL)
“I am so appreciative to have learned and experienced so many new things related to open source.”
-Vaibhav Chopra, PyBaMM, Benchmarks
“I learned a lot about the Julia programming language, surrogate models, automatic differentiation and so much more. My mentors were incredibly supportive and I found the feedback that they provided on my code to be extremely enriching.”
-Vikram Narayan, SciML, Improve Surrogates.jl
“GSOC was a brilliant opportunity to contribute to the open source community. I was able to learn from both my mentors and others in the space, as well as dive deep into the theory and application of projection predictive inference.”
-Yann McLatchie, PyMC, Implement projection predictive model selection for Bambi-fitted GLMs
“Working with the open source community is a super joy.”
-Yilin Xia, ArviZ, ArviZ Dashboards
“I have learned many things about C++ programming while contributing to the community and had a deeper understanding of parallel computing and pipeline models. I really appreciate that I can work with my mentor, who taught me many exciting and useful skills.”
-Zhicheng Xiong, Taskflow, Enhance Taskflow’s Pipeline Infrastructure