variance and standard deviation are imprecise when size is not high #418
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
In my opinion the current variance (and by consequence the standard deviation) computation are imprecise, when curSize is not high (when curSize < 10 for example):
```java
variance = (sumSquares / curSize) - (mean * mean);
```
The computation I suggest in this pull request does not hurt and is precise:
```java
if (curSize == 1) {
variance = 0d;
} else {
variance = (sumSquares - ((double) total * total / curSize)) / (curSize - 1);
}
```
For reference:
http://web.archive.org/web/20050512031826/http://helios.bto.ed.ac.uk/bto/statistics/tress3.html
brharrington
approved these changes
Jun 4, 2017
There was a problem hiding this comment.
Thanks for the contribution. Looks good to me.
FYI, we do not use StatsTimer much internally. The computed stats are for a local node and do not work very well when slicing and dicing by various dimensions or looking at aggregates across the cluster which are the most common use-cases for us.
The basic timers report 4 stats: total, totalOfSquares, count, and max. The first three allow us to compute std dev across an arbitrary grouping by computing the sum of those stats and then using them to compute std dev server side.
| } | ||
| mean = (double) total / curSize; | ||
| variance = (sumSquares / curSize) - (mean * mean); | ||
| if (curSize == 1) { |
There was a problem hiding this comment.
It would be good to add a test case for current size of 1. At first glance none of the existing tests will check this branch.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
In my opinion the current variance (and by consequence the standard deviation) computation is imprecise, when curSize is not high (when curSize < 10 for example):
The computation I suggest in this pull request does not hurt and is precise:
For reference:
http://web.archive.org/web/20050512031826/http://helios.bto.ed.ac.uk/bto/statistics/tress3.html