
Sunset Sunrise grid aggregation & chi
The notebooks are stored as markdown files with jupytext for better git compatibility.
These notebooks can be run with jupyterlab-docker.
All figures (png, pdf, interactive html, svg) available in the resources folder.
Notebooks (ipynb) can be found in the notebooks folder.
Results
In addition to the resource folder, the latest HTML converts of notebooks are also available here:
- 01_grid_agg.html
- Results: 16 static maps for Instagram and Flickr (Sunrise + Sunset) and Flickr Totals,
absolute User Count, Post Count and User Days
- 02_visualization.html
- Results: Interactive Visualization using Bokeh, with most popular Flickr CC Images shown on hover
- Example outputs:
- 03_chimaps.html
- Results: Chi maps for Flickr and Instagram Sunset and Sunrise, for User Count and User Days, and
for Natural Breaks and Head Tail Breaks classification schemes, excluding non-significant chi values
- Outputs:
- with non-significant:
- without non-significant:
- User Count:
- User Days:
- Post Count:
- 04_combine.html
- Results: Merged chi values (positive), to combine Instagram and Flickr results for comparison
- Outputs:
- Head Tail Breaks:
- Natural Breaks:
- 05_countries.html
- Results: Country aggregate chi values, also for producing relationship plots
- Outputs:
- Head Tail Breaks:
- Natural Breaks:
- Quantiles:
- 06_semantics.html
- 07_time.html
- Temporal aggregate data per month, for Flickr and Instagram
- 08_relationships.html
- Relationship plots for comparison:
- Flickr Sunrise + Flickr Sunset
- Instagram Sunrise + Instagram Sunset (Flickr expected)
- Sunrise Flickr + Sunrise Instagram
- Sunset Flickr + Sunset Instagram
- and a test with Instagram Sunset+Sunrise as expected
- For comparison, we also created these relationship plots
for the chi values for each country: Relationship chi plots for comparison.
Albeit interesting, there was not enough space in the paper to discuss these results.
- 09_statistics.html
- Total aggregates, Summary Statistics
Among other outputs, there are two core result maps produced in notebooks:
Convert to ipynb files
First, either download release files or convert the markdown files to working jupyter notebooks.
To convert jupytext markdown files to ipynb-format:
If you’re using the docker image, open a terminal inside jupyter and follow these commands:
bash
conda activate jupyter_env && cd /home/jovyan/work/
Afterwards, re-create the .ipynb
notebook(s) with:
mkdir notebooks
jupytext --set-formats notebooks///ipynb,md///md,py///_/.py --sync md/01_grid_agg.md
jupytext --set-formats notebooks///ipynb,md///md,py///_/.py --sync md/02_visualization.md
jupytext --set-formats notebooks///ipynb,md///md,py///_/.py --sync md/03_chimaps.md
jupytext --set-formats notebooks///ipynb,md///md,py///_/.py --sync md/04_combine.md