Coevolving Innovations

… in Business Organizations and Information Technologies

Learning data science, hands-on

For the Quantitative Methodologies for Design Research (定量研究方法) course for Ph.D. students at Tongji University in spring 2017, Susu Nousala invited me to join the team of instructors in collaborative education in Shanghai.  Experts were brought in during the course to guide the graduate students.

My participation in the course over two days had three parts:  (a) preparing a lecture outline; (b) orienting the students; and (c) equipping the students with tools.

(A) Preparing a lecture outline

While I’m comfortable with the mathematics underlying statistical analysis, I have a lot of practical experience of working with business executives who aren’t.  Thus, my approach to working with data relies a lot on presentation graphics to defog the phenomena.  While the label of data science began to rise circa 2012, I’ve had the benefit of practical experience that predates that.

Today's APL
AGSS: A Graphical Statistical System (1994)

In my first professional assignment in IBM Canada in 1985, data science would have been called econometrics.  My work included forecasting country sales, based on price-performance indexes (from the mainframe, midrange and personal computer product divisions) and economic outlooks from Statistics Canada.  Two years before the Macintosh II would bring color to personal computing, I was an early adopter of GRAFSTAT: “An APL system for interactive scientific-engineering graphics and data analysis” developed at IBM Research.  This would eventually become an IBM program product by called AGSS (A Graphical Statistical System) by 1994.

Metaphor Computer Systems workstation
Metaphor Computer Systems workstation

In 1988, I had an assignment where data science would have been called marketing science.  I was sent to California to work in the IBM partnership with Metaphor Computer Systems. This was a Xerox PARC spin-off with a vision that predated the first web page on the World Wide Web by a few years.  These activities led me into the TIMS Marketing Science Conference in 1990, cofounding the Canadian Centre for Marketing Information Technologies (C2MIT) and contributing chapters to The Marketing Information Revolution published in 1994.

This journey led me to appreciate the selection and use of computer-based tools for quantitative analysis.  Today, the two leading platforms in “Data Science 101” are Python (a general purpose language with statistical libraries), and the R Project for Statistical Computing (a specialized package for data analysis and visualization).  Both are open source projects, and free to download and use on personal computers.  I tried both.  R is a higher level programming language more similar to the APL programming language that gets work done more quickly.  For statistical work, I recommend R over Python (although APL is a theoretically better implementation).

Intro to R Programming, Big Data University
Intro to R Programming, Big Data University, Feb. 22, 2017

Since I live in Toronto, I attended the February session of Data Science with R – Bootcamp in person, at Ryerson University.  There, I was watched Polong Lin leading a class through R using the Jupyter notebook, both in (i) an interactive version, and (ii) a printable version.  Students had the choice to either follow Polong (i) actively, in a step-by-step execution in the Cognitive Class Virtual Lab (formerly called the Data Scientist Workbench) with a cloud-based R session through their web browsers, or (ii) passively, reading the static printable content.

Polong was helpful in guiding us with course resources that would be available in Shanghai.  In North America, we have cognitiveclass.ai with course materials, and datascientistworkbench.com as a cloud computing platform (that includes R and Python).  For China, there are parallel sites at bigdatauniversity.com.cn and datascientistworkbench.cn in Chinese, that native speakers could be more comfortable using.

Based on Polong’s materials, I developed a Jupyter notebook additionally emphasizing graphical presentation for an in-person lecture in Shanghai.  These materials are cached at coevolving.com/tongji/201704_DataScienceR/, where they are accessible globally.

(B) Orienting the students

Ph.D. students in design will not have been required to have studied mathematics at the university level.  However, I then recalled that Shanghai high school students had a history with PISA achievement of “top of the global class in maths with an average score … or the equivalent of nearly three years of schooling, above the average” in 2013.

Ph.D. course in Quantitative Methods, Tongji University College of Design and Innovation
This strength flavoured the description of my approach for the lecture to students:

  • Did you study linear algebra?  (Pause to have a Chinese professor translate that.)
  • Of course, you studied linear algebra in high school.
  • So, how do I approach exploratory data analysis?
  • I plot the data, and draw a straight line through it.
  • If the plot doesn’t look right, I move the data so that straight line looks right.

An Orientation Demonstration (Jupyter notebook)

The workshop scope was then explained as:

  • NOT to teach you everything about data science!
  • You should know about ways to represent data to support your research findings.
  • You will see some good tools for qualitative methods (and maybe even quantitative methods).
  • We can work together to get you started on tools that will suit your needs.

Working with the end in mind, the orientation stepped through a recommended package with three tools:  (i) ggplot2, through (ii) Jupyter, on (iii) R.

The demonstration loaded an example of Housing, with Sales Prices of Houses in the City of Windsor as 546 rows and 12 variables.  This showed that a higher housing price correlates with a larger lot size, something that makes sense intuitively.  The ggplot2 library has a nice feature of facet grids, so that incorporating the number of bedrooms into an analysis can be visualized as a collection of plots (rather than a single plot).

(C) Equipping the students with tools

Having then shown the students what outputs could look like, we released them to experiment as a self-study group, with the professors available as standby resources.  As Ph.D. students at one of the top universities in China, they were expected to step up to the challenge.  Since neither Susu nor I are proficient in Mandarin, the students would likely learn faster as a group if not slowed down by second-language conversion.

The students then enthusiastically set upon getting their tools to work.  They collectively surfaced two challenges:  (i) technical; and (ii) research use.

(i) The technical issue was that, although the bigdatauniversity.com.cn web site was readily accessible, the students found the cloud-based datascientistworkbench.cn slow over the Internet.  While we in North America expect high bandwidth speeds in our workplaces and homes, the responsiveness through the browser interface at Tongji University to the cloud platform was too slow.  This led the students to prefer installing open source software tools onto their personal computers, dissolving Internet connectivity problems.

Most of the students were running MacBooks, and a few were on Windows laptops.  Downloading R from the Comprehensive R Archive Network and installing that program was easy.  Downloading Anaconda from Continuum Analytics as a step towards a Jupyter notebook was relatively straightforward.  But then, with both Anaconda and R up and running, getting Jupyter to connect to R was a challenge.  This required opening up a Terminal — a new experience for most on MacOS, but somewhat more familiar as command line on Windows — and typing in magic incantations.

  • The installation of IRKernel is not a drag-and-drop activity, and sometimes requires more technical knowledge to interpret the messages signalling other than success.
  • MacOS X with Jupyter needs some Python libraries that requires the installation of XCode.
  • MacOS X after 10.5 (Leopard) no longer includes X11, requiring a separate installation of XQuartz.
  • On MacOS X, installing the IRKernel would not work if R was launched by clicking on the R icon, but would if R was started from a Terminal.

Resolving these issues had the Ph.D. students collectively helping each other for 6 hours in the afternoon and evening on the first course day, and then 2 hours on the second morning.  Updating MacOS X with XCode over the Internet was slow on the university network.  On Windows, one student easily got all of the packages working on his computer, and then struggled to help a friend struggling with similar hardware and software.

The ordeal of installing software had an unexpected benefit of becoming a team-building activity.  The students banded together in mutual support!

(ii) The research use issue arose as the students worked their way through the Big Data University exercises.  Exercises typically use datasets already prepackaged for use,  so students can focus on the programming language.  Real research projects requiring bringing in real datasets.  Pointing out the import features for R programming moved the students one step ahead.

In a course oriented towards a future of big data, this raised the question:  where is the data?  The phenomenon of open data — as electronic data sources readily accessible over the Internet — is nascent within the People’s Republic of China.  This led to questions about research design, where students would have to determine questions of interest, and plan ways to collect data.

After those two days, my participation in the course diminished with a few exchanges over WeChat.  Susu brought some of the students onto a research project in Southern China, putting their research design learning into practice over the summer.  By the fall, the students should have performed preliminary data analysis, so that coaching may be welcomed on finer points about statistics.


Leave a Reply

Your email address will not be published. Required fields are marked *

  • Recent Posts

  • Archives

  • RSS qoto.org/@daviding (Mastodon)

    • New status by daviding April 21, 2019
      Public libraries can become hubs for peer-to-peer learning. In the Let's Learn Teach Online program, #TorontoLibrary has partnered with #P2PU, #CiscoNetAcad, #TorontoESS, and #GBCollege to facilitate "Linux Unhatched" and "Introduction to IoT". Larysa Essex shared their experiences at the @gtalug meeting on April 9, 2019. https://daviding.wordpress.com/2019/04/20/2019-04-09-larysa-essex-linux-unhatched-learning-circles-at-toronto-public-library-web-video/
    • New status by daviding April 1, 2019
      Afternoon break in 200-year-old mid-lake pavilion included zhong, quail eggs, kumquats, sesame peanut blocks, preserved plums. Following afternoon visiting two art museums, the snack re-energized us into discussing philosophy, following the tradition of those frequenting Chinese teahouses. (Yuyuan Tea House, Yu Garden, Shanghai, PR China) 20190331 @marcocataffo
    • New status by daviding April 1, 2019
      Here in Shanghai, @marcocataffo has a Thinkpad T430 , which I've now brought up to date with Manjaro Linux (and Kubuntu LTS as a backup) alongside Windows 7. He's now 2 days jet lagged from Italy. Eventually, maybe @antlerboy will meet somewhere.
    • daviding shared a status by antlerboy@mastodon.social February 9, 2019
      @daviding Wittgenstein:"6.54 My propositions are elucidatory in this way: he who understands me finally recognizes them as senseless, when he has climbed out through them, on them, over them. (He must so to speak throw away the ladder, after he has climbed up on it.)"
    • New status by daviding February 9, 2019
      Dinner with @rms @fsf inviting the activists #CivicTechTO to gain some insight into discussions on privacy concerns #QuaysideToronto. We outlined but didn't delved into the complexity of three levels of government involved in #WaterfrontTO. (Royal Myanmar, Homer Avenue, Etobicoke, Ontario) 20190208
  • RSS on IngBrief

    • Contextual dyadic thinking (Lee, 2017)
      Contextual dyadic thinking is proposed by Keekok Lee in her 2017 The Philosophical Foundations of Classical Chinese Medicine. This is as a way of appreciating Chinese implicit logic, as an alternative to dualistic thinking that has developed over centuries in Western philosophy.
    • Dao, de, wei, wuwei (Lai 2003)
      Appreciating wei and wuwei has led to the context of dao and de, in the writings of Karyn L. Lai. The scholarly review acknowledges prior interpretations of de and dao.
    • Engineering Resilience vs. Ecological Resilience (Holling, 1996)
      For @theNASciences in 1996, #CSHolling clarified definitions of resilience, with engineering seeking one equilibrium state, while ecology recognizes many. Those who emphasize the near-equilibrium definition of engineering resilience, for example, draw predominantly from traditions of deductive mathematical theory (Pimm,. 1984) where simplified, untouched ecological systems are imagined, or from traditions of engineering, where the motive […]
    • Service coproductions as reciprocal activities
      In addition to extrinsic economic exchange, #JohnMCarroll #JiaweiChen #ChienWenTinaYuan #BenjaminHanrahan @ISTatPENNSTATE say service coproductions relying on all participants to collaborate in both economic exchange and social exchange. Service coproduction is a special case of service provision in which the roles of service provider and service recipient both require active participation. Examples include healthcare, education, and […]
    • Science and Society in East and West | Joseph Needham | 2004
      In researching #SystemsChange, fundamental differences in science and philosophy in the west and the Chinese were surfaced by #JosephNeedham. A useful translation of wéi and wú wéi (i.e. 為 and 無為 , or 为 and 无为) is the ways of "human will" and "nature" as juxtaposed.
    • Wiki as computational platform
      Thinking forward on #federatedwiki, rather than backwards by @wardcunningham. > [Federated wiki] is a computational platform for the collaborative construction of things that work and will continue to work as platform technology evolves underneath it. > Too much thinking about wiki as a note-taking system will just hold it back.
  • RSS on Media Queue

  • RSS on daviding.com

    • 2019/03 Moments March 2019
      Month of intensive lectures and research meetings, in Toronto and then in Shanghai, with social breaks on local excursions to clear minds.
    • 2019/02 Moments February 2019
      Reduced exercise outside with a cold and snowy February, with excursions out of the house to warm places with family, friends and colleagues.
    • 2019/01 Moments January 2019
      January in Toronto has lots of intellectual offerings and artistic exhibitions to attract the curious out of warm homes, through cold and snow.
    • 2018/12 Moments December 2018
      Tried to have a normal month, with a busy social calendar of birthdays, a funeral plus Christmas season, while daily temperatures hovered just above freezing.
    • 2018/11 Moments November 2011
      Mentally busy month with a conference coming to town, and maintaining the regular pattern of local meetings, travel around town only by bicycle.
    • 2018/10 Moments October 2018
      October had more bicycling cross-town as fall temperatures declined, plus a 6-day trip to Portland Oregon for pattern language conferences.
  • Meta

  • Creative Commons License
    This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
    Theme modified from DevDmBootstrap4 by Danny Machal