Note: JupyterDay Philly will be preceded on May 17-18 by Bryn Mawr’s sixth annual Blended Learning in the Liberal Arts Conference, a forum for faculty and academic support staff to share their experiences with incorporating digital technologies into teaching. Jupyter Day attendees are welcome to come, please see the link above for more details.

Jupyter Workshop and Tutorials

Thursday, May 18, 2017 – Maneesha Sane, workshop details

JupyterDay Philly Schedule

Friday, May 19, 2017 in Dalton Hall at Bryn Mawr College.

  • Registration and Light Breakfast, coffee – 8:00 – 9:00
  • Welcome 9:00 – 9:15, Douglas Blank, Bryn Mawr College
  • Keynote talk –  9:15 – 10:15
    • Andrew Odewahn, CTO of O’Reilly Media- [PAPER]
      Lowering the Barriers to Entry to Jupyter in Academia: Lessons from the Publishing Industry
      Jupyter Notebooks and other Open Source tools provide a huge opportunity for educators (particularly in STEM, but in other domains, as well) to create new and engaging types of content that go far beyond static text. However, because they’re often built by hackers for hackers, using these tools effectively in a course often feels more like writing software than a course. So while the benefits are huge, reaping the rewards often requires a steep learning curve. For almost 5 years, O’Reilly Media has centered its publishing processes around tools like Jupyter, git, GitHub, Docker, and a host of Open Source packages. In this talk, O’Reilly CTO will talk about what has worked (and what hasn’t!) in trying to deploy these tools, and provide practical guidance on how to use them in an educational setting.
  • Coffee break 10:15 – 10:30
  • Sessions 10:30-12:00, Moderator Liz McCormack
    • Maneesha Sane, Data Carpentry and Software Carpentry – [SLIDES]
      Teaching Python: Command line or Jupyter Notebook?
      Python is one of the most popular languages for new programmers to learn. It can be taught using many different platforms and environments, including the Jupyter Notebook, the command line interface, and other IDEs as teaching tools. All environments present unique opportunities and challenges to learning and teaching. In this talk I will explore the differences specifically between the Jupyter notebook and the command line interface. I will examine how an instructor’s motivations to teach, students’ motivations to learn, and the context in which teaching, learning, and working happens can influence the tools we use to teach with.
    • Joshua Shapiro, Bryn Mawr College, Biology – [SLIDES]
      Combining Introductory Biology and Computer Science with Jupyter
      As biology enters the post-genomic era, the increasingly rapid accumulation of biological information has led to ever further integration of the field with computer science. Students of biology increasingly encounter computational approaches in the biological literature, and students of computer science are ever more likely to encounter biological applications of the concepts they are studying. We designed a course to allow students with interests in computer science and biology to explore both fields simultaneously, preparing them for more advanced courses in either field or, ideally, in both. The course was run almost entirely through the use of Jupyter Notebooks and JupyterHub, including lectures with live coding, active learning exercises, and assignments. Through a mixture of simulations and data analysis, students were introduced to computational topics including types, functions, recursion, and computational efficiency, exploring biological concepts in ecology, evolutionary biology, and genomics. I will present ways in which course took advantage of the Jupyter platform to meet our broad goals, describe areas of success, and discuss improvements planned for future iterations of the course.
    • Steven Neshyba, University of Puget Sound, Chemistry
      What (and how) do students learn when they use Jupyter Notebooks in the classroom?
      Computational Guided Inquiry (CGI) is an in-class, student-centered teaching method in which students engage in guided inquiry within a computational environment. Assessment of the efficacy of the approach is still in early stages, however. One key set of questions surrounds our limited understanding of the way students learn in such an environment. For example, at what point in a CGI session do students achieve key learning milestones? What, if anything, is fundamentally different about how students retain those learning milestones? Here, I present assessment instruments aimed at improving our understanding of how students learn in a CGI session. The instruments are not directed at learning outcomes, but rather specifically focused on student perceptions of their own learning processes. Students are asked to rank the various activities during a given CGI session (instructor exposition, non-computational guided inquiry, and guided inquiry) in order of how effective those elements were in advancing their learning, and how much that learning was perceived to be under their own control (vs instructor control). Time during the talk will be set aside to discuss the possible merits and pitfalls of such instruments, and perspectives of workshop participants on other instruments.
    • Douglas Blank, Bryn Mawr College, Computer Science – [SLIDES]
      Computational Storytelling in the Liberal Arts
      In this talk I explore the use of notebooks in the liberal arts as a means of telling stories. In this manner, notebooks can elevate computation to the level of reading and writing, making it part of the core tools of communication. But this goal is not without challenges. From use in computer science courses to firstyear writing seminars, notebook-based computing is useful, rewarding, and addresses issues in equity. But we will need to refine both the notebook and our teaching styles in order to gain the most benefit.
  • Lunch 12:00 – 12:45 pm (buffet  lunch provided, Dalton 212a, e)
  • Demonstrations – 12:45 – 1:45 (Canaday 315)
  • Keynote talk – 2:00 – 3:00
    • Luis Martí, Institute of Computing, Universidade Federal Fluminense – [SLIDES]
      Jupyter Notebooks in a Computational Intelligence/Machine Learning Class
      Jupyter notebooks are a language-agnostic computational framework that encapsulates in a single platform-independent interactive document source code, text, graphics, and other media. Because of these features, together with the emergence of a related dynamic software ecosystem, notebooks have established themselves as one of the tools for choice for documenting and sharing experiments and support materials. Naturally, Jupyter notebooks have found their way into the classroom. Their use has frequently been related to programming courses or as support tools in practical/seminar classes. However, their features make them a suitable pedagogical tool for presenting complex topics, like those related to computational intelligence and machine learning. In this talk, we debate the different approaches on how to apply jupyter notebooks for teaching computational intelligence and machine learning. We will provide different examples of classes and how jupyter notebooks allow to enhance the learning experience of students.
  • Break 3:00 – 3:15
  • Sessions 3:15 – 4:45, Moderator Doug Blank
    • Aaron Titus, High Point University – [SLIDES]
      Teaching Physics with Computation Using Jupyter Notebook and VPython
      Departmental learning objectives in physics at High Point University include theory, experimental physics, and computational modeling. All three are important, and all three are woven into the four year curriculum. Physics students learn VPython in the introductory calculus-based physics course using “Matter and Interactions” by Ruth Chabay and Bruce Sherwood. Simultaneously, physics majors engage in a year-long introductory research project that often includes both experimental and computational components. This approach prepares students to use computational modeling as a tool for research and problem solving, starting in their first year. Furthermore, it has an extraordinary impact on growing a thriving department. As students gain experience with computation, they transition from GlowScript VPython to Jupyter. I will provide details of our approach, examples of student work, and evidence of its impact.
    • Mark Matlin, Bryn Mawr College, Physics – [SLIDES]
      Teaching Computational Skills in the Natural Sciences using Jupyter Notebooks in a Blended Approach
      For several years, the Physics Department at Bryn Mawr College has tried to find a way to teach scientific computing skills to our students without requiring an additional course to complete the major, and without needing additional staffing.  With the support of a grant from the Helmsley Foundation administered by the Association of American Colleges & Universities’ TIDES (Teaching to Increase Diversity and Equity in STEM) program, we have developed a suite of modular lessons on scientific computing using Jupyter notebooks.  These modules are meant to be used either in courses or as standalone learning resources.  Besides teaching computing skills, the collection of modules and associated materials are intended to promote student metacognition and illuminate the diversity of the community of individuals who use computation in their work.  In this talk I will describe the development and current status of this suite of modules, as well as its implementation in our curriculum, and share thoughts on possible future directions for the suite.
    • Gunjan Baid, University of California – [SLIDES]
      Data Science at UC Berkeley: 2000 undergraduates, 50 majors, no command line
      UC Berkeley’s Data Science education program has no math, computing, or statistics prerequisites and is designed to be accessible to students of all backgrounds. At the introductory level, the program consists of Data 8, a fundamentals course that introduces students to concepts of computer programming and statistics. Alongside Data 8, there is a diverse set of connector courses that allow students to apply data science to their area of interest, such as geography, immunotherapy, or cognitive science. Using Jupyter notebooks, students are able to get hands-on experience working with data without the burden of setting up and maintaining a development environment. For these courses, we have developed various tool to further simplify the student workflow. Students can obtain notebooks and datasets for an assignment with one click. Autograding, user authentication, and submission are also all done through Jupyter notebooks. Over 2000 students across 50 majors have taken our fundamentals course and the connector courses in the past four semesters. In this presentation, I will explain our program in more detail and expand upon the pedagogical challenges we have faced in scaling Jupyter notebooks for use in large courses. I will then discuss how our tools and content can be applied to teaching Data Science using Jupyter at other universities and institutions.
    • Jacob Frias Koehler, New York University, Mathematics – [SLIDES]
      What about PyCalc?  An experimental PreCalculus course for Undergraduates
      What role might scientific computing play for students pursuing non-STEM majors?  This talk describes an experimental precalculus course in which students used Jupyter notebooks with Python as the primary problem-solving tool.  Here, students from non-technical majors with no prior experience computing worked to develop and refine a series of Jupyter notebooks that were subsequently shared through a public github repository.  The resulting shifts in mathematical content and student activity are covered, giving way to students’ narratives and exemplars of work.
  • Wrap up 4:45 – 5:00, Douglas Blank