Prerequisites:
- Recent R environment (tested on 3.2.5).
- Internet connection (for installation of extra packages).
- Compiler toolchain and development libraries (for building R packages, distribution-dependent).
Usage:
./analyze-overlap.r
The command above should generate two files:
output-diagram.pdf: chord diagram illustrating overlap between data sourcesoutput-stats.csv: table with overlap statistics for the analyzed sources
Default input file:
anonymized-overlap-data-2015-07--2016-06.csv
Input format:
Tab-separated columns, no headers, no quotes. Columns:
- First: name of the analyzed source.
- Second: name of another source that shares an attribute. Can be null (\N) if the attribute is not shared with other sources.
- Third: value of the attribute shared between the sources.
Example:
vendor12 alexa 8.8.8.8
vendor12 cuckoo 8.8.4.4
vendor12 cuckoo 10.0.0.1
vendor12 cuckoo 10.0.0.2
vendor12 \N 10.0.0.3
...
IP addresses and selected source names in the sample input file have been anonymized.
For more details, read the source.