Analysis of parameters

[peterdesmet]

I would visualize the parameters as a stacked bar chart. We'll need 4:

• [x] Number of datasets
• [x] Number of occurrences
• [x] Number of datasets (GBIF context)
• [x] Number of occurrences (GBIF context)

  Bar chart format

Each stacked bar chart has this format:

parameter	true	false	?
use	green	red	gray
distribution	green	red	gray
derivatives	green	red	gray
commercial	green	red	gray
attribution	red	green	gray
share alike	red	green	gray
notification	red	green	gray

• The values are either number of datasets or number of occurrences.
• Since the values are stacked, each bar represents the total and the parts the percentages.
• Hoovering over a part ideally shows the total and percentage.
• For the colours we could choose: one colour per type (true, false, ?) or - like I did above - a colour based on open/restricted.

  GBIF context

Since most datasets don't have a license, interpreting what that means (see issue #12) has a huge influence on the results. Therefore, we also need to analyze the data where [no license] is interpreted as:

standard license : 
use : true
distribution: true
derivatives : true
commercial : ?
attribution : true
share alike : ?
notification : false

[peterdesmet]

It looks like chart.js does not support stacked bar charts yet. There's a pull request with this feature though.

[bartaelterman]

Done for datasets: https://github.com/Datafable/gbif-data-licenses/blob/analyse-licenses/output/parameters-distribution.html

[bartaelterman]

And an update of that file now also contains visualisations on the occurrence level. This raises some visual issues (tooltip position, y-axis position). I'll raise separate issues for those.

[bartaelterman]

I suggest we discuss the GBIF context later.

[peterdesmet]

Tooltip issue solved (#22).

[peterdesmet]

For GBIF context, interpret [no license] as described above and join back with datasets as a file called datasets-annotated-gbif-dua.

[bartaelterman]

Code added in #25 to interpret the [no license] as described above + generate visualisations.