Memory Blossoms

Data: Phenological data series of cherry tree flowering in Kyoto, Japan, from 800 AD to 2021, collected by Yasuyuki Aono, Keiko Kazui, and Shizuka Saito

Blog: Here is my blog post that explains the design process of the project.

Contents:

• Data-viz Drawing
• Data-viz with D3.js
• Statistical Thinking in Python

• Visit this notebook and create your own Memory Blossoms!
• Ask for help! If anyone has any suggestions on how to create petals without changing some data points with dummy data, please let me know :) Any advice would be appreciaated.

Data-viz Drawing

I was inspired to begin my data visualization studies through "Data-viz drawing" as presented by designer Giorgia Lupi and Stefanie Posavec in Dear Data. Participating in online Data-viz drawing workshop with Stefanie Posavec on Jun 2-3, 2021 and crafting something imperfect on my own, I try to deepen my understanding of "data visualization" that goes beyond basic charts, on top of what I learned at the IIT Institute of Design's Data Visualization workshop Fall 2020.

• Visit this page to learn more about messy(ish) experimentation.
  

Visual Inspiration

Breaking down a subject into visual variables

Lovely visuals by Staphanie Posavec

Color Gradients

• Python code left, right
• Reference left, right
  

Variations of shapes of petals

• See SVG code here.
• Reference here.

Data-viz with D3.js

Resources

Highly recommend those free online tutorials for D3.js + JS beginners.

• D3 in Depth
• D3 for R Users
• D3.js Playbook
• freeCodeCamp

Also, visit my collection of Intro to D3.js @Observable

Graphical Exploratory Data Analysis

• Some notes I've taken at the online dataviz workshop by Shirley Wu

Scatterplot

Observable notebook here.

Area chart with Missing Data

Observable notebook here.

Line chart with Missing Data

Observable notebook here.

Radial line chart with Missing Data

Observable notebook here.

Histogram

Observable notebook here.

Statistical Thinking in Python

• Reference: Statistical Thinking in Python (Part 1/2) @datacamp
• Note

Special thanks to #データ可視化の学び場, @hayataka88, and everyone in the community!

Graphical Exploratory Data Analysis

Histogram

See Python code here.

Scatter plot

Color represents peak-bloom date. Size represents temperature. See Python code here.

Quantitative Exploratory Data Analysis

Comparing percentiles to ECDF

percentiles: [401. 409. 414. 418. 426.] float64

See Python code here.

Box-and-whisker plot

See Python code here.

Thinking Probabilistically

Linear regression

slope = -0.015823592193534804 estimated temp / peak bloom date
intercept = 14.114277800365738 estimated temp

See Python code here.

Visualizing bootstrap samples

mean: 409.42857142857144
median: 409.0
std: 3.917516914514882

See Python code here.

Bootstrap replicates

sem: 0.5783480621330582
std: 0.5818109818422588
95% confidence interval = [408.24489796 410.51020408]

See Python code here.

95% confidence interval for the mean

See Python code here.