Exploring exploratory programming

TL;DR: In this article, an in-depth interview with five code artists with diverse backgrounds, levels of experience, and working on different code-based artistic expressions was conducted, identifying characteristics and commonalities in their development process, the tools they use, their sources of inspiration, and their expectations.

Abstract: In this study, we examine the work of data scientists, members of an emerging technical occupation, through the lens of craft. Drawing on 65 in-depth interviews with data scientists, we show that their work cannot be adequately explained by the human–machine configurations characterized in the existing literature on craft in technical occupations, which primarily focuses on crafting products using ready-to-use tools and ready-to-be-processed materials. Instead, we find that data scientists craft not only their products, but also their tools and materials, often in an iterative and non-linear fashion. This distinct approach entails a unique human–machine-data configuration that we refer to as data-based craft, which stems from the unique nature of digital data and learning algorithms that data scientists simultaneously craft and use. This study advances our understanding of craft in the digital age by highlighting the need to reconceptualize human–machine relationships in data-intensive occupations.

TL;DR: A novel way to provide a tangible, consolidated notion of both space and time in object exploration tools to make it more likely that programmers will use the available means to explore the evolution of particular objects is presented.

TL;DR: TRACTUS is a system extending the popular RStudio IDE, that detects, tracks, and visualizes code experiments in hypothesis-driven data science tasks, and helps recall decisions and insights by grouping code experiments into hypotheses, and structuring information like code execution output and documentation.

TL;DR: This work describes new algorithms that take into account the variable cost of learning algorithm experiments and that can leverage the presence of multiple cores for parallel experimentation and shows that these proposed algorithms improve on previous automatic procedures and can reach or surpass human expert-level optimization for many algorithms.

TL;DR: A description of some typical EDA procedures is given and some of the principles of EDA are discussed.

TL;DR: "Digital fluency" should mean designing, creating, and remixing, not just browsing, chatting, and interacting.