What's Wrong with Computational Notebooks? Pain Points, Needs, and Design Opportunities

Chattopadhyay, Sankha; Prasad, Ishita; Henley, Austin Z.; Sarma, Anita; Barik, Titus

doi:10.1145/3313831.3376729

Cited by 121 publications

(101 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These approaches are motivated by recognizing that neither direct manipulation nor coding is best suited for all tasks, but combining them yields an accumulation of benefits -users can rapidly and intuitively specify designs via direct manipulation, but then switch to code to construct reusable abstractions. This goal resonates with the results of recent surveys and interviews of data scientists which find that visual interfaces are most useful if their output can be captured in code [12,17,57]. Thus, akin to Sketch-N-Sketch and Wrangler, B2 provides bidirectional bridges between code and interactive visualizations: B2 synthesizes appropriate visualizations by tracing the data lineage expressed in code, and interactions performed on the visualizations are logged to code cells to enable further analysis.…”

Section: Interactions Generating Codementioning

confidence: 63%

“…In contrast, B2 records interaction histories as predicates, a representation that is tailored to interactive visualization but is also more robust to bidirectional changes. Moreover, by taking a more focused scope, B2 identifies and addresses an additional challenge with integrating code and visualizations that may not apply to graphical interfaces more generally: restricting interactive visualizations to a linear flow of interleaved cell outputs limits the creation of richer multi-view coordination [17,56].…”

Section: Interactions Generating Codementioning

confidence: 99%

“…B2's goal of bridging code and interactive visualization is motivated by recent surveys and interviews of data scientists [12,14,17,57]. In particular, Wongsuphasawat et al find that data scientists often switch between several tools including textual environments (e.g., MATLAB or Jupyter) and graphical interfaces (e.g., Tableau or Microsoft PowerBI) during their analysis sessions [57].…”

Section: The Needs Of Data Scientistsmentioning

confidence: 99%

“…In particular, Wongsuphasawat et al find that data scientists often switch between several tools including textual environments (e.g., MATLAB or Jupyter) and graphical interfaces (e.g., Tableau or Microsoft PowerBI) during their analysis sessions [57]. As Chattopadhyay et al report, this switching behavior forces analysts to repeat themselves by manually translating work they conducted in code to visual interfaces, or vice-versa [17]. For many data scientists, this overhead is sufficiently prohibitive that they eschew visual analysis tools altogether and restrict themselves to working only in code [57].…”

Section: The Needs Of Data Scientistsmentioning

confidence: 99%

See 3 more Smart Citations

B2

Hellerstein

Satyanarayan

2020

Proceedings of the 33rd Annual ACM Symposium on User Interface Software and Technology

View full text Add to dashboard Cite

Data scientists have embraced computational notebooks to author analysis code and accompanying visualizations within a single document. Currently, although these media may be interleaved, they remain siloed: interactive visualizations must be manually specified as they are divorced from the analysis provenance expressed via dataframes, while code cells have no access to users' interactions with visualizations, and hence no way to operate on the results of interaction. To bridge this divide, we present B2, a set of techniques grounded in treating data queries as a shared representation between the code and interactive visualizations. B2 instruments data frames to track the queries expressed in code and synthesize corresponding visualizations. These visualizations are displayed in a dashboard to facilitate interactive analysis. When an interaction occurs, B2 reifies it as a data query and generates a history log in a new code cell. Subsequent cells can use this log to further analyze interaction results and, when marked as reactive, to ensure that code is automatically recomputed when new interaction occurs. In an evaluative study with data scientists, we find that B2 promotes a tighter feedback loop between coding and interacting with visualizations. All participants frequently moved from code to visualization and vice-versa, which facilitated their exploratory data analysis in the notebook.

show abstract

Section: Interactions Generating Codementioning

confidence: 63%

Section: Interactions Generating Codementioning

confidence: 99%

Section: The Needs Of Data Scientistsmentioning

confidence: 99%

Section: The Needs Of Data Scientistsmentioning

confidence: 99%

See 2 more Smart Citations

B2

Hellerstein

Satyanarayan

2020

Proceedings of the 33rd Annual ACM Symposium on User Interface Software and Technology

View full text Add to dashboard Cite

show abstract

“…Version control for notebooks has been the focus of several notebook plugins, many blog posts and feature requests (e.g., jupyterlab-git 3 , verdant 4 , nbdime 5 ), and several research studies [23,34,51,53]. Some notebook services like Nextjournal 6 value the importance of preserving history in computational notebooks so much, that they offer automatic versioning of the notebook and related artifacts.…”

Section: Use Of Version Controlmentioning

confidence: 99%

Code Duplication and Reuse in Jupyter Notebooks

Andreas

Ernst

Storey

2020

2020 IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC)

View full text Add to dashboard Cite

Reusing code can expedite software creation, analysis and exploration of data. Expediency can be particularly valuable for users of computational notebooks, where duplication allows them to quickly test hypotheses and iterate over data, without creating code from scratch. In this thesis, I'll explore the topic of code duplication and the behaviour of code reuse for Jupyter notebooks; quantifying and describing snippets of code and explore potential barriers for reuse. As part of this thesis I conducted two studies into Jupyter notebooks use. In my first study, I mined GitHub repositories, quantifying and describing code duplicates contained within repositories that contained at least one Jupyter notebook. For my second study, I conducted an observational user study using a contextual inquiry, where my participants solved specific tasks using notebooks, while I observed and took notes. The work in this thesis can be categorized as exploratory, since both my studies were aimed at generating hypotheses for which further studies can build upon. My contributions with this thesis is twofold: a thorough description of code duplicates contained within GitHub repositories and an exploration of the behaviour behind code reuse in Jupyter notebooks. It is my desire that others can build upon this work to provide new tools, addressing some of the issues outlined in this thesis.

show abstract

Data science knowledge integration: Affordances of a computational cognitive apprenticeship on student conceptual understanding

Sánchez‐Peña

Vieira

Magana

2022

Comp Applic In Engineering

View full text Add to dashboard Cite

This study implements a computational cognitive apprenticeship framework for knowledge integration of Data Science (DS) concepts delivered via computational notebooks. This study also explores students' conceptual understanding of the unsupervised Machine Learning algorithm of K‐means after being exposed to this method. The learning of DS methods and techniques has become paramount for the new generations of undergraduate engineering students. However, little is known about effective strategies to support student learning of DS and machine learning (ML) algorithms. The research questions are: How do students conceptualize their understanding of an unsupervised ML method after engaging with interactive visualizations designed using the computational cognitive apprenticeship approach? How do the affordances of the interactive visualizations support or hinder student knowledge integration of an unsupervised machine learning method? Design‐based research allowed for the iterative design, implementation, and validation of the pedagogy in the context of a working classroom. For this, data collection methods often take the form of student artifacts. We performed a qualitative content analysis of students' written responses and reflections elicited during the learning process. Results suggest that the computational cognitive apprenticeship promoted knowledge integration. After interacting with the computational notebooks, most students had accurate conceptions of the goal and the nature of the method and identified factors affecting the output of the algorithm. Students found it useful to have a concrete representation of the method, which supported its conceptual understanding and showcased the acquisition of strategic knowledge for its appropriate execution. However, we also identified important misconceptions students held about the algorithm.

show abstract

What's Wrong with Computational Notebooks? Pain Points, Needs, and Design Opportunities

Cited by 121 publications

References 31 publications

B2

B2

Code Duplication and Reuse in Jupyter Notebooks

Data science knowledge integration: Affordances of a computational cognitive apprenticeship on student conceptual understanding

Contact Info

Product

Resources

About