This data set is my personal log of running reviews in Open Exchange.
It is just a table with 555 records and no special features.
Ready to run experiments on various ways of indexing.
Personal names of authors are scrambled for this example.
All data are resulting from the analysis of the OEX web pages.
The first run collects the URL for the Details pages skipping duplicates
imposed by the "featured" header block.
The next run collects updates and details of the packages as available.
Some informatinon is missing as it gets inserted dynamically over several
fetch cycles while I have just 1 from IRIS.
A utility to load and update is included. To use these methods you will
need to create an SSL Configuration named "community" in SMP
for client access to community.intersystems.com:443 .
This is included in Docker image setup.
Make sure you have git and Docker desktop installed.
Clone/git pull the repo into any local directory
git https://github.com/rcemper/Dataset-OEX-reviews.git
Run the IRIS container with your project:
docker-compose up -d --build
Connect to the containers SMP and examine content in namespace USER
applying the decribed examples or use commandline $system.SQL.Shell()
or use Online Demo :
SELECT count(id) pkg, stars
FROM dc_data_rcc.OEX
group by NVL(stars,-1)
order by 2 desc
----------------------------------------------
pkg stars
7 6.0
1 5.5
66 5.0
24 4.5
14 4.0
13 3.5
14 3.0
6 2.5
9 2.0
4 1.5
4 1.0
5 0.5
337 0.0
----------------------------------------------
SELECT stars, %exact(author) author
FROM dc_data_rcc.OEX
where stars > 5 group by author order by 1 desc
----------------------------------------------
stars author
6.0 ZVPUNRY-O_NNZ
6.0 `UdUb0cdUYgUb
6.0 N\PSSH\TL'YVUNPLY
6.0 TY]O*^YWK]*]KV`KNY\
6.0 X[^QZf[,_OMXQ_Q
6.0 ZN_V\-`N[PURg-ZNPVN`
6.0 fXeZXl3`\^[T\_Xa^b
5.5 J[LJS^%XM[FWT[
5.5 HQMXV]$QEWPIRRMOSZ
5.5 rnkb9fZkq9i^k^bkZ9`hf^l
----------------------------------------------
Logo added
Initial Release