[WIP] Paper about the UDF interface

[uellue]

This is very much work in progress. Posted here because of previous positive experience with collaborative writing: #460

Possible target journal: https://www.journals.elsevier.com/information-systems (focus on data intensive applications, scientific computing)

Points to consider:

• Make clear WHY we are building this interface in the first place.
• Go beyond electron microscopy
• Compare API with Hadoop, Spark, Dask arrays (anything else?)
  • How to do that objectively?
  • Which criteria? Performance, ease of use, scalability, flexibility, ...?
• Show clearly why the API is what it is, reasons for its structure and features
• Show where it is innovative and how it can be of general use beyond being an implementation detail of LiberTEM
• Balance UDF and Blobfinder aspects
• Show how Blobfinder is the lead application, the application is much wider, however
• Keep it short and precise, not a boring reference documentation
• Don't require insider knowledge of LiberTEM or electron microscopy to understand
• Cover both the design process as well as the result on high, scientific level

In that sense, feedback and suggestions for changes, including sweeping ones, are very welcome!

Contributor Checklist:

• I have added myself to the creators.json file

[codecov]

Codecov Report

Merging #570 into master will decrease coverage by 0.02%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #570      +/-   ##
==========================================
- Coverage   58.81%   58.79%   -0.03%     
==========================================
  Files         208      208              
  Lines        8316     8316              
  Branches     1062     1062              
==========================================
- Hits         4891     4889       -2     
- Misses       3229     3230       +1     
- Partials      196      197       +1

Impacted Files	Coverage Δ
src/libertem/udf/blobfinder/correlation.py	97.99% <0%> (-0.81%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update bd55311...5546c79. Read the comment docs.