Repository navigation
Create error estimator for SUM - #528
Conversation
| ratio_dropped_linf = self.get_ratio_dropped_linf(linf_bound) | ||
| ratio_dropped = 1 - (1 - ratio_dropped_l0) * (1 - ratio_dropped_linf) | ||
| stddev = self._get_stddev(l0_bound, linf_bound) | ||
| if to_print: |
There was a problem hiding this comment.
is it for debugging only? should we remove it? just double checking :)
There was a problem hiding this comment.
Yes, thanks. Removed
| delta: Optional[float] = None, | ||
| ): | ||
| return histogram_error_estimator.create_error_estimator( | ||
| return histogram_error_estimator.create_estimator_for_count_privacy_id_count( |
There was a problem hiding this comment.
I would rename from _get_estimator to _get_estimator_for_count_and_privacy_id_count.
|
|
||
| @parameterized.parameters((0, 1), (0.5, 0.89), (1, 0.78), (2, 0.76), | ||
| (40, 0)) | ||
| # there 1 is contribution 40 and 10 contribution 1. |
There was a problem hiding this comment.
there is 1 contribution of 40 and 10 contributions of 1.
| (40, 0)) | ||
| # there 1 is contribution 40 and 10 contribution 1. | ||
| # total contribution = 1*40+10*1 = 50 | ||
| # when linf_bound = 0.5, left after contribution bounding 11*0.5=5.5, i.e. |
There was a problem hiding this comment.
how linf can be a double and not an integer?
There was a problem hiding this comment.
linf_bound is max contribution per partition, which means
max_contributions_per_partition for COUNT
max_sum_per_partition for SUM (which can be double)
dvadym
left a comment
dvadym
left a comment
There was a problem hiding this comment.
Thanks for review!
| ratio_dropped_linf = self.get_ratio_dropped_linf(linf_bound) | ||
| ratio_dropped = 1 - (1 - ratio_dropped_l0) * (1 - ratio_dropped_linf) | ||
| stddev = self._get_stddev(l0_bound, linf_bound) | ||
| if to_print: |
There was a problem hiding this comment.
Yes, thanks. Removed
| delta: Optional[float] = None, | ||
| ): | ||
| return histogram_error_estimator.create_error_estimator( | ||
| return histogram_error_estimator.create_estimator_for_count_privacy_id_count( |
|
|
||
| @parameterized.parameters((0, 1), (0.5, 0.89), (1, 0.78), (2, 0.76), | ||
| (40, 0)) | ||
| # there 1 is contribution 40 and 10 contribution 1. |
| (40, 0)) | ||
| # there 1 is contribution 40 and 10 contribution 1. | ||
| # total contribution = 1*40+10*1 = 50 | ||
| # when linf_bound = 0.5, left after contribution bounding 11*0.5=5.5, i.e. |
There was a problem hiding this comment.
linf_bound is max contribution per partition, which means
max_contributions_per_partition for COUNT
max_sum_per_partition for SUM (which can be double)
2 participants
ErrorEstimatorusesContributionHistogramsto estimate approximately RMSE error for different L0/Linf bounds. This PR allows to createErrorEstimatorfor SUM (the different with COUNT, only different set of histograms)