Skip to content

Commit 6a9967d

Browse files
committed
Making Figures
1 parent 4abdb65 commit 6a9967d

3 files changed

Lines changed: 48 additions & 86 deletions

File tree

_guides/01_lab.md

Lines changed: 2 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -23,14 +23,9 @@ title: Navigating Lab Meetings
2323
## Frequency and Duration
2424

2525
### Frequency
26-
- Lab meetings are held *weekly* at the same day and time (TBD)
26+
- Lab meetings are held *weekly* at the same day and time (Monday 10:30)
2727
- Lab meetings are held in-person in room 4105 at the Western Interdisciplinary Research Building unless specified otherwise.
28-
- *Note*: No lab meetings are held on holidays
2928

30-
### Time Limit
31-
- Lab meetings should take no longer than *one hour*.
32-
33-
---
3429

3530
## Lab Meeting Formats
3631

@@ -41,9 +36,8 @@ title: Navigating Lab Meetings
4136

4237
- If you are presenting, make sure to have your setup fully ready before the start of the lab meeting. This includes setting up hybrid presentations if needed.
4338

44-
4539
### Tutorials/workshops
46-
- List of potential tutorials TBD
40+
- We often give tutorials of specific topics. These should be relevant to at least a number of lab members. Even if they are not immediately relevant to your work, they are a great way to learn something new. If you think there is a specific topic we should have a tutorial on, please don't hesitate to suggest the topic. We will then find somebody who can run the tutorial.
4741

4842
### Journal clubs
4943
- A team member selects a scientific article for a group discussion. The presenter summarizes the article's objectives, mthods and results followed by a group discussion.

_guides/07_making_figures.md

Lines changed: 0 additions & 78 deletions
This file was deleted.

_guides/20_making_figures.md

Lines changed: 46 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,46 @@
1+
---
2+
layout: page
3+
title: Making awesome figures
4+
usemathjax: true
5+
---
6+
A large part of human brain is devoted to image processing. So it should not be a big surprise that many (human) scientists mostly look at Figures and captions.
7+
When designing figures for a paper, the goal is to produce something that is so compelling that the reader can grasp the main points of the paper without even reading a single word.
8+
Also, figures should be aesthetically pleasing and have style.
9+
10+
# Stage 1: Plan each figure panel
11+
## Determine the communicative goal
12+
The main function of each figure panel is to communicated a point (or multiple points) between the writer and the reader. When readers look at a well-designed figure panel, they immediately understand what the point is - without having to read the caption or paper. These points can help us to get close to this level of clarity.
13+
14+
To achieve this, you need to be very clear for yourself which points you want to convey with each panel. If you, as the figure-maker cannot verbalize the goal of your figure, then your figures are doomed to be messy, and convoluted. Note that you cannot drive home all the points in your paper in one single graph. Sometimes less is more. Do not occlude data with data.
15+
16+
## Sort the points
17+
Each of your desired points can be conveyed with a visual contrast (chance in color, shape, or location of ink). Some of the visual contrasts are easier to perceive so pay attention to how you choose the association. Space > color > shape
18+
### Minimalism
19+
It does not matter if you are looking at a machine, a short story, or a scientific figure. In any system, when you know every part has a role, and there is no redundancy, it gives a special kind of satisfaction. You appreciate the design, and the thoughtfull designer. So make sure, you use the minimum amount of ink. Do not use unnecassary ink in your plot unless there is an essential role for it to play. With each line in the graph, ask the question of whether it conveys any information. If not, delete it - the figure will likely look better without. Typical examples include: Boxes around graph panels, boxes around legends, and grid-lines are typical graph-junk that should be removed.
20+
21+
22+
# Stage 2: Fine tune and assemble graphs into good-looking figures
23+
Once you figure out the critical points and have decided which figure panels you need to tell your story, it is time to bring the figures to something that is publication worthy. There is a big difference between a good-enough figure, and one that impresses with elegance and style. These step can help you make your figures look not only professional but aesthetically pleasing to look at.
24+
25+
### Assemble figure panels into figure and scale to journal size
26+
For the final assembly of a figure, use a vector-based graphics program. Affinity Designer, Illustrator, Inkspace, are good choices. Power point is not.
27+
For each figure then decide how wide your figure should be in the final pdf document of your paper and make a document of the correct size.
28+
For journal papers, common width are:
29+
* 1 column: 8.5cm
30+
* 1.5 columns: 11.6cm
31+
* 2 columns: 17.6 cm
32+
33+
Then try to arrange your panels to fit the width. Ensure that you have a adequate spacing of the panels without too much space between them.
34+
Once you have done this with all Figures, you need to ensure that the formatting is consistent across all figure panels and figures.
35+
36+
- Stroke width should be harmonious - at journal size, I recommend 0.75pt for figure axes, and 1-2pt for lines indicating data
37+
- Use one single font across all figures (A good sans-serif font as Arial or Myriad Pro is a good choice). Don't mix different fonts, unless you have a special communicative goal to do so.
38+
- At journal size, no font should be smaller than 6pt
39+
- As a general advice, the following font sizes look good:
40+
- Subpanel letter (a,b,): 18-25
41+
- Figure titles: 10-12pt
42+
- Axis Main Label: 9-11pt
43+
- Tick labels: 7-9pt
44+
- Legend entrees 8-10pt
45+
46+

0 commit comments

Comments
 (0)