AlexanderSaydakov commented on a change in pull request #107:
URL:
https://github.com/apache/datasketches-website/pull/107#discussion_r759577364
##########
File path: docs/Theta/ThetaSetOpsCornerCases.md
##########
@@ -82,60 +79,38 @@ We can assign a single octal digit ID to each state where
The octal digit `ID = ((theta == 1.0) ? 4 : 0) | ((retainedEntries > 0) ? 2 :
0) | (empty ? 1 : 0);`
-| Shorthand Notation | theta | retained entries | empty |
Has Seen Data | ID | Comments |
-|:---------------------------------:|:-----:|:----------------:|:----------:|:-------------:|:--:|:------------------------------:|
-| New {1.0,0,T} | 1.0 | 0 | T |
F | 5 | New Sketch, p=1.0 (default) |
-| Exact {1.0,>0,F} | 1.0 | >0 | F |
T | 6 | Exact Mode |
-| Estimation {<1.0,>0,F} | <1.0 | >0 | F |
T | 2 | Estimation Mode |
-| NewDegen {<1.0,0,T}<sup>2</sup> | <1.0 | 0 | T |
F | 1 | New Sketch, user sets p<1.0 |
-| ResultDegen {<1.0,0,F}<sup>3</sup>| <1.0 | 0 | F |
T | 0 | Valid Intersect or AnotB result |
+| Shorthand Notation | theta | retained entries | empty |
Has Seen Data | ID | Comments |
+|:---------------------------------:|:-----:|:----------------:|:----------:|:-------------:|:--:|:-------------------------------:|
+| Empty {1.0,0,T} | 1.0 | 0 | T |
F | 5 | Empty Sketch |
+| Exact {1.0,>0,F} | 1.0 | >0 | F |
T | 6 | Exact Mode |
+| Estimation {<1.0,>0,F} | <1.0 | >0 | F |
T | 2 | Estimation Mode |
+| Degenerate {<1.0,0,F}<sup>3</sup> | <1.0 | 0 | F |
T | 0 | Valid Intersect or AnotB result |
---
-<sup>2</sup> *New Degenerate*: New Empty Sketch where the user sets *p < 1.0*.
This can be safely reinterpreted as {1.0,0,T} because it has not seen any
data.<br>
-<sup>3</sup> *Result Degenerate*: Can appear as a result of a an Intersection
or AnotB of certain combination of sketches.
-
-## Invalid States of a Sketch
-The remaining three combinations of the variables represent internal errors
and should not occur.
Review comment:
OK, I restored this section. The "empty flag" column in the table was
wrong, so I corrected it. I am not convinced this is needed. It does not really
explain _why_. These are just invalid states. One of them, which was the
NewDegen before, is used in practice as the initial state of the update sketch
with p < 1, but this is an implementation detail, and is not visible from the
outside. I am not sure we need to explain this, and it was a part of my
motivation to remove this section. If it is not very useful, I would call it
clutter.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
