- CSS 95%
- Python 5%
Warming Gradient #
CSS gradient visualizing historical temperature data, inspired by Ed Hawkins' Warming Stripes.
Installation #
You can use NPM (best for projects with a build process), a CDN (quick copy-paste into any HTML file), or download and include the CSS file yourself.
Option 1: Install with NPM #
Make sure you have Node.js installed and run this command in your project folder:
npm install warming-gradient
Import in CSS #
Add this at the top of your main CSS file:
@import 'node_modules/warming-gradient/warming-gradient.css';
Or just paste this <link> in the <head> of your HTML file:
<link rel="stylesheet" href="node_modules/warming-gradient/warming-gradient.css" />
The CSS class and variable will now be available in your project.
Option 2: Use via CDN #
Just paste this <link> in the <head> of your HTML file and the CSS class and variable will be available in your project:
<link rel="stylesheet" href="https://unpkg.com/warming-gradient" />
Option 3: Self-host warming-gradient.css #
- Download the
warming-gradient.cssfile from the source files here. - Place it in your project folder.
- Include it in your HTML:
<link rel="stylesheet" href="path/to/warming-gradient.css" />
How to use #
Apply the class or variable directly in your HTML, for example:
<!-- Use with class -->
<div class="warming-gradient-global" style="height: 100px"></div>
<!-- Use with variable -->
<div style="height: 100px; background: var(--warming-gradient-global)"></div>
Regional Variables #
For regional data, CSS variables follow the pattern --warming-gradient-{region} (e.g., --warming-gradient-europe-netherlands). Use these for specific regions:
<!-- Use with class -->
<div class="warming-gradient-europe-netherlands" style="height: 100px"></div>
<!-- Use with variable -->
<div style="height: 100px; background: var(--warming-gradient-europe-netherlands)"></div>
Contributing #
Data Structure #
Organize datasets in a folder structure with each dataset in its own subfolder containing a settings.json:
data/
europe/
netherlands/
netherlands.csv
settings.json
germany/
germany.txt
settings.json
global/
global.csv
settings.json
Generating CSS #
Run the batch processor to generate CSS for all datasets:
python3 generate_all.py data/ -e 2025
This creates:
warming-gradient.css- Combined CSS with all gradientssettings_log.json- Log of all settings and CSS variable names
CSS Variable Naming #
Each dataset's CSS variable follows the pattern --warming-gradient-{path}, where {path} is the folder path from the data directory with hyphens instead of slashes. For example:
data/europe/netherlands/→--warming-gradient-europe-netherlandsdata/global/global/→--warming-gradient-global
Use the variable in your CSS:
.my-element {
background: var(--warming-gradient-europe-netherlands);
}
Settings #
Each dataset's settings.json supports:
| Key | Description | Default |
|---|---|---|
name |
CSS name override (without prefix) | Folder path (e.g., europe-netherlands) |
file |
Data filename | Only .csv/.tsv/.txt file in folder |
begin |
First year | 1900 (local) or first year (global) |
end |
Last year | Last complete year |
scale |
Color scale (std dev multiple) | 3.0 (local) |
fixed_scale |
Color scale (degrees C) | 0.9 (global) |
read |
Data reading options (see below) | auto-detected |
source |
Data source note (not used) | - |
Read options:
| Key | Description | Default |
|---|---|---|
header_line |
Header row number | First non-blank line |
year_column |
Year column name/number | Column 1 |
value_column |
Temperature column | Column 2 |
delimiter |
Field separator | Detected (comma, semicolon, tab) |
decimal |
Decimal mark | . |
missing |
Missing value indicators | [] |
type |
Data type | local or global (auto-detected) |
Example settings.json:
{
"begin": 1864,
"scale": 3.0,
"read": {
"header_line": 16,
"value_column": "year",
"missing": ["NA"]
},
"source": "MeteoSwiss"
}
Command-Line Options #
| Flag | Description |
|---|---|
-s, --scale |
Color scale in standard deviations |
-f, --fixed-scale |
Color scale in degrees C |
-b, --begin |
Begin year |
-e, --end |
End year |
-o, --output-dir |
Output directory |
--css-name |
Output CSS filename |
Command-line flags override settings.json for all datasets.
How It Works #
Color mapping #
- Blue shades: Below baseline (cooler temperatures)
- Red shades: Above baseline (warmer temperatures)
- Gray: Missing data
Reference periods #
- Baseline (1961-2010): Defines the midpoint where blue transitions to red
- Standard Deviation (1901-2000): Used to scale color intensity (only if a custom scale is provided)
Scaling modes #
- Fixed scale: Uses a constant value (default 0.9 for global, 3.0 std dev for local)
- Calculated scale: Multiplies the calculated standard deviation by your specified value
Read my blog post for more information
License #
Code licensed under MIT license. Please double-check the license for all data yourself.