search

andrewbutterfield / CS4098-2017

1 stars 1 forks source link

Language Constructs and DDIs Sample Data #2

Open CiaranCostello opened 7 years ago

CiaranCostello commented 7 years ago

The readme in the examples folder says there is a ConstructDDITest.csv file. This isn't there.

The results given for Identify Sequential DDIs seem incorrect. I would have thought that there would also be DDIs (2,3) (2,4) (3,4) in SEQ1, if there are DDIs between all of the drugs.

These DDIs would depend on the missing csv file. Could we get a csv file to clear it up?

CiaranCostello commented 7 years ago

I was wrong, the csv file is in the main directory.

It does have a DDI between (2,3) (2,4) and (3,4). Should these not be reported in the sequential results for SEQ 1?

andrewbutterfield commented 7 years ago

Its in the folder above! I'll move it

On 6 Apr 2017, at 21:47, CiaranCostello notifications@github.com wrote:

The readme in the examples folder says there is a ConstructDDITest.csv file. This isn't there.

The results given for Identify Sequential DDIs seem incorrect. I would have thought that there would also be DDIs (2,3) (2,4) (3,4) in SEQ1, if there are DDIs between all of the drugs.

These DDIs would depend on the missing csv file. Could we get a csv file to clear it up?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/andrewbutterfield/CS4098-2017/issues/2, or mute the thread https://github.com/notifications/unsubscribe-auth/AAj59yramdRZk4xp7mGZuWY89MeoCkbUks5rtU9FgaJpZM4M2Jc6.

Andrew Butterfield School of Computer Science & Statistics Trinity College Dublin 2, Ireland

andrewbutterfield commented 7 years ago

yes SEQ1 should have a DDI between 2,4 and 3,4 but not 2,3 - that should be reported for SEQ2 - the closest enclosing sequence construct