chart-tufte

Name: chart-tufte
Author: chacosoldier

29stars6forks

How to install chart-tufte?

chacosoldier/compabob/chart-tufte

$npx -y skills add chacosoldier/compabob --skill chart-tufte

Installs into the current project.

Summary

Self-grade rubric for any quantitative chart, grounded in Edward Tufte's *Visual Display of Quantitative Information*. Run as a final pass inside visual-explainer (or any chart-emitting workflow) before showing the chart to the user. Strips chartjunk, checks lie factor, picks t

Files · 1

View on GitHub

SKILL.md

1	# Chart Tufte
2
3	A chart is good when it shows the data, helps the viewer reason about it, and does not lie. Tufte spent a career formalising this. This skill collapses his principles into a self-grade pass that runs on any chart the assistant is about to emit.
4
5	Read [references/vdqi-catalogue.md](references/vdqi-catalogue.md) first when grading. It holds the named-failures library (NYT MPG 14.8, TIME barrel 59.4, etc.) and the named-exemplars library (Minard, Marey, Playfair, Snow). The catalogue is what makes assessment diagnostic instead of generic.
6
7	## When to invoke
8
9	- Automatically — as the final step inside `visual-explainer` whenever the output is a quantitative chart (bar, line, scatter, area, dot, range-frame). Skip for diagrams (architecture, sequence, flow); those are different beasts.
10	- Explicitly — `/chart-tufte <chart-spec>` to score a chart produced elsewhere.
11
12	## The rubric (run before emitting)
13
14	Score each on 0-10. Stop and revise if any score drops below 5.
15
16	1. Data-ink ratio — does every pixel of ink represent data? Borders, gridlines, redundant legends are non-data ink. Target: tend toward 1.0; typical for default-styled charts is 0.1-0.2 (VDQI p.136).
17	2. Lie factor — `(visual change %) / (data change %)`. Acceptable range 0.95-1.05. Cite a VDQI named failure if it is worse: TIME's barrel chart hit 59.4, NYT's MPG hit 14.8.
18	3. Data density — numbers per square inch. Below 0.15 is overwrought (a single bar showing one number is the canonical sin). High-density alternatives: small multiples, range-frame scatters, tables.
19	4. Chartjunk count — does the chart contain any of the four named species?
20	- Moiré — vibrating cross-hatch patterns.
21	- The dreaded grid — gridlines drawn darker than the data.
22	- The duck — visual gimmick that overwhelms the data (3D pies, gradient-shaded bars).
23	- Decoration — clip art, icons, mascots in the chart frame.
24	Each chartjunk species costs points.
25	5. Genre fit — is this the right shape for this data?
26	- ≤20 numbers → table, not chart
27	- Many series → small multiples (one panel per series)
28	- 2 quantitative variables → range-frame scatter (axis only spans data min-max)
29	- Distribution → quartile plot, not bar of mean
30	- Time series → thin line + direct endpoint labels, no legend
31	6. Dimensionality discipline — does the chart use more dimensions than the data has? A 3D pie chart on 1D data is dishonest by construction. Penalise.
32	7. Direct labelling — are series labelled at their endpoint or inline, not via a separate legend? Legends force the reader's eye to bounce.
33	8. Range-frame discipline — does the axis span only the data range, not 0 to some arbitrary max?
34	9. Comparability — if multiple panels: same y-scale where comparison matters; different scales only when local shape is the question.
35
36	## Remedies (when the rubric flags problems)
37
38	- B1 — fix lie factor: redraw with proportional scaling. State the new lie factor in the chart caption.
39	- B2 — erase non-data ink: remove borders, drop gridlines, kill the legend if direct labelling fits.
40	- B3 — increase data density: switch to small multiples if you have ≥3 panels of comparable data.
41	- B4 — pick the right genre (see C1-C10 below).
42	- B5 — name the failure: if the chart resembles a VDQI named failure (NYT MPG 14.8, TIME barrel 59.4, LA Times shrinking-doctor 2.8, USA Today 3D bar chart, Playfair's wheat decline), say so before showing.
43	- B6 — fix dimensionality: 1D data → 1D chart; never 3D unless the third dimension is in the data.
44	- B7 — deflate currency over time: if plotting dollars across years, deflate to a common year (state which) before plotting. Failing to do this is a silent lie factor.
45
46	## Genre playbook (C1-C10)
47
48	- C1 — quartile plot for distributions (5-number summary, no whiskers gimmick).
49	- C2 — range-frame scatter for two quantitative variables (axes span data range only).
50	- C3 — dot-dash marginals on a range frame to show density on each axis.
51	- C4 — paired bars for two-group comparison on one categorical axis.
52	- C5 — small multiples for many series (1 panel per series, same x-axis, often same y-axis).
53	- C6 — sparkline for inline time series at word-scale.
54	- C7 — slopegraph for before/after comparison across many categories.
55	- C8 — stem-and-leaf table when the goal is the actual numbers, not the shape.
56	- C9 — table when ≤20 numbers and structure matters.
57	- C10 — thin line chart for time series, direct endpoint label, no legend, no gridlines.
58
59	## Exemplars to emulat

People who liked this also viewed

6 picks

Type

chart-tufte

29stars6forks

View on GitHub

How to install chart-tufte?

chacosoldier/compabob/chart-tufte

$npx -y skills add chacosoldier/compabob --skill chart-tufte

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Chart Tufte
2
3	A chart is good when it shows the data, helps the viewer reason about it, and does not lie. Tufte spent a career formalising this. This skill collapses his principles into a self-grade pass that runs on any chart the assistant is about to emit.
4
5	Read [references/vdqi-catalogue.md](references/vdqi-catalogue.md) first when grading. It holds the named-failures library (NYT MPG 14.8, TIME barrel 59.4, etc.) and the named-exemplars library (Minard, Marey, Playfair, Snow). The catalogue is what makes assessment diagnostic instead of generic.
6
7	## When to invoke
8
9	- Automatically — as the final step inside `visual-explainer` whenever the output is a quantitative chart (bar, line, scatter, area, dot, range-frame). Skip for diagrams (architecture, sequence, flow); those are different beasts.
10	- Explicitly — `/chart-tufte <chart-spec>` to score a chart produced elsewhere.
11
12	## The rubric (run before emitting)
13
14	Score each on 0-10. Stop and revise if any score drops below 5.
15
16	1. Data-ink ratio — does every pixel of ink represent data? Borders, gridlines, redundant legends are non-data ink. Target: tend toward 1.0; typical for default-styled charts is 0.1-0.2 (VDQI p.136).
17	2. Lie factor — `(visual change %) / (data change %)`. Acceptable range 0.95-1.05. Cite a VDQI named failure if it is worse: TIME's barrel chart hit 59.4, NYT's MPG hit 14.8.
18	3. Data density — numbers per square inch. Below 0.15 is overwrought (a single bar showing one number is the canonical sin). High-density alternatives: small multiples, range-frame scatters, tables.
19	4. Chartjunk count — does the chart contain any of the four named species?
20	- Moiré — vibrating cross-hatch patterns.
21	- The dreaded grid — gridlines drawn darker than the data.
22	- The duck — visual gimmick that overwhelms the data (3D pies, gradient-shaded bars).
23	- Decoration — clip art, icons, mascots in the chart frame.
24	Each chartjunk species costs points.
25	5. Genre fit — is this the right shape for this data?
26	- ≤20 numbers → table, not chart
27	- Many series → small multiples (one panel per series)
28	- 2 quantitative variables → range-frame scatter (axis only spans data min-max)
29	- Distribution → quartile plot, not bar of mean
30	- Time series → thin line + direct endpoint labels, no legend
31	6. Dimensionality discipline — does the chart use more dimensions than the data has? A 3D pie chart on 1D data is dishonest by construction. Penalise.
32	7. Direct labelling — are series labelled at their endpoint or inline, not via a separate legend? Legends force the reader's eye to bounce.
33	8. Range-frame discipline — does the axis span only the data range, not 0 to some arbitrary max?
34	9. Comparability — if multiple panels: same y-scale where comparison matters; different scales only when local shape is the question.
35
36	## Remedies (when the rubric flags problems)
37
38	- B1 — fix lie factor: redraw with proportional scaling. State the new lie factor in the chart caption.
39	- B2 — erase non-data ink: remove borders, drop gridlines, kill the legend if direct labelling fits.
40	- B3 — increase data density: switch to small multiples if you have ≥3 panels of comparable data.
41	- B4 — pick the right genre (see C1-C10 below).
42	- B5 — name the failure: if the chart resembles a VDQI named failure (NYT MPG 14.8, TIME barrel 59.4, LA Times shrinking-doctor 2.8, USA Today 3D bar chart, Playfair's wheat decline), say so before showing.
43	- B6 — fix dimensionality: 1D data → 1D chart; never 3D unless the third dimension is in the data.
44	- B7 — deflate currency over time: if plotting dollars across years, deflate to a common year (state which) before plotting. Failing to do this is a silent lie factor.
45
46	## Genre playbook (C1-C10)
47
48	- C1 — quartile plot for distributions (5-number summary, no whiskers gimmick).
49	- C2 — range-frame scatter for two quantitative variables (axes span data range only).
50	- C3 — dot-dash marginals on a range frame to show density on each axis.
51	- C4 — paired bars for two-group comparison on one categorical axis.
52	- C5 — small multiples for many series (1 panel per series, same x-axis, often same y-axis).
53	- C6 — sparkline for inline time series at word-scale.
54	- C7 — slopegraph for before/after comparison across many categories.
55	- C8 — stem-and-leaf table when the goal is the actual numbers, not the shape.
56	- C9 — table when ≤20 numbers and structure matters.
57	- C10 — thin line chart for time series, direct endpoint label, no legend, no gridlines.
58
59	## Exemplars to emulat

People who liked this also viewed

6 picks

Type