Fork It: Supporting Stateful Alternatives in Computational Notebooks

Weinman, Nathaniel; Drucker, Steven M.; Barik, Titus; DeLine, Robert

doi:10.1145/3411764.3445527

Cited by 36 publications

(22 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To address this issue, researchers introduce techniques to clean unused code [5,19] and provide better version control of code [7,15]. To align notebook designs with the non-linear and iterative nature of exploratory data analysis, Weinman et al [22] explore alternatives to the single execution state of notebooks with forking and backtracking, as well as a two-column layout. Similarly, researchers have also explored designs that allow users to more easily navigate between code and interactive visualizations, such as bi-directional communications [10] and side-by-side presentations [24].…”

Section: Background and Related Workmentioning

confidence: 99%

“…Since there is only one code interpreter that keeps all the execution states in a notebook, users need to carefully arrange and execute cells in certain orders to avoid variable corruptions [5,8]. In addition, all notebooks follow a linear presentation style that is contradictory to the non-linear nature of exploratory data analysis [9,22]. Effectively, all of the code cells are a part of a single program, despite being spread across the notebook.…”

Section: Introductionmentioning

confidence: 99%

“…This means that edits such as changing variables in one part of the notebook may have unforeseen consequences later in the notebook. To help users better organize notebook cells, researchers propose techniques such as distilling essential cells to create sub-notebooks [5], supporting cell-level code version controls [15], and two-column notebook layout [22]. In contrast to these works, we address the notebook messiness by focusing on the linearity of the traditional notebook layout.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

StickyLand: Breaking the Linear Presentation of Computational Notebooks

Wang,

Dai,

Edwards

2022

Preprint

View full text Add to dashboard Cite

Figure 1: The StickyLand user interface persists on top of a computational notebook, creating an always-on workspace for data scientists to display text and execute code. (A) A user can drag an existing Notebook Cell to (B) the Sticky Dock that mounts on the edge of the notebook. (C) It creates a Sticky Cell that stays at the same location even when the user scrolls the notebook. (D) The user can further convert this cell to a Floating Cell that floats on top of the notebook with customizable position and size.

show abstract

Section: Background and Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

StickyLand: Breaking the Linear Presentation of Computational Notebooks

Wang,

Dai,

Edwards

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…For example, messes in code may accrue during EDA, and data scientists may lose track of their thought processes. To address these issues, several tools have been proposed, such as Variolite [27], Verdant [28], and ForkIt [53], to support fast versioning and history tracking. Code Gathering Tools [19] assist data scientists with cleaning, recovering, and comparing versions of code in notebooks by analyzing code cells' dependency and organizing code into chunks.…”

Section: Computational Notebooksmentioning

confidence: 99%

EDAssistant: Supporting Exploratory Data Analysis in Computational Notebooks with In-Situ Code Search and Recommendation

Li¹,

Zhang²,

Leung³

et al. 2021

Preprint

View full text Add to dashboard Cite

Using computational notebooks (e.g., Jupyter Notebook), data scientists rationalize their exploratory data analysis (EDA) based on their prior experience and external knowledge such as online examples. For novices or data scientists who lack specific knowledge about the dataset or problem to investigate, effectively obtaining and understanding the external information is critical to carry out EDA. This paper presents EDAssistant, a JupyterLab extension that supports EDA with in-situ search of example notebooks and recommendation of useful APIs, powered by novel interactive visualization of search results. The code search and recommendation are enabled by state-of-the-art machine learning models, trained on a large corpus of EDA notebooks collected online. A user study is conducted to investigate both EDAssistant and data scientists' current practice (i.e., using external search engines). The results demonstrate the effectiveness and usefulness of EDAssistant, and participants appreciated its smooth and in-context support of EDA. We also report several design implications regarding code recommendation tools.

show abstract

“…This might happen when, for example, removing outliers or experimenting with different hyper-parameters. Thus, research has looked at supporting such experimental workflows in order to manage versions of notebooks [9,23], and support cleanup and refactoring [8]. Related to that is work that uses tools to debug data flows in large (non-notebook) data analytics pipelines [16], or support transparency in statistical reporting with multiverse analysis [6].…”

Section: Introductionmentioning

confidence: 99%

Error Identification Strategies for Python Jupyter Notebooks

Robinson,

Ernst,

Vargas

et al. 2022

Preprint

View full text Add to dashboard Cite

Computational notebooks-such as Jupyter or Colab-combine text and data analysis code. They have become ubiquitous in the world of data science and exploratory data analysis. Since these notebooks present a different programming paradigm than conventional IDE-driven programming, it is plausible that debugging in computational notebooks might also be different. More specifically, since creating notebooks blends domain knowledge, statistical analysis, and programming, the ways in which notebook users find and fix errors in these different forms might be different. In this paper, we present an exploratory, observational study into how notebook users find and understand potential errors in notebooks. We presented users with notebooks pre-populated with common notebook errors-errors rooted in either the statistical data analysis, the knowledge of domain concepts, or in the programming. We then analyzed the strategies our study participants used to find these errors and determined how successful each strategy was at identifying errors. Our findings indicate that while the notebook programming environment is different from the environments used for traditional programming, debugging strategies remain quite similar. It is our hope that the insights presented in this paper will help both notebook tool designers and educators make changes to improve how data scientists discover errors more easily in the notebooks they write.

show abstract

Fork It: Supporting Stateful Alternatives in Computational Notebooks

Cited by 36 publications

References 23 publications

StickyLand: Breaking the Linear Presentation of Computational Notebooks

StickyLand: Breaking the Linear Presentation of Computational Notebooks

EDAssistant: Supporting Exploratory Data Analysis in Computational Notebooks with In-Situ Code Search and Recommendation

Error Identification Strategies for Python Jupyter Notebooks

Contact Info

Product

Resources

About