Expand description
This crate provides histogram implementations that are conceptually similar to HdrHistogram, with modifications to the bucket construction and indexing algorithms that we believe provide a simpler implementation and more efficient runtime compared to the reference implementation of HdrHistogram.
§Goals
- simple implementation
- fine-grained configuration
- efficient runtime
§Background
Please see: https://observablehq.com/@iopsystems/h2histogram
Structs§
- A histogram that uses atomic 64bit counters for each bucket.
- A bucket represents a quantized range of values and a count of observations that fall into that range.
- The configuration of a histogram which determines the bucketing strategy and therefore the relative error and memory utilization of a histogram.
- A histogram that uses plain 64bit counters for each bucket.
- This histogram is a sparse, columnar representation of the regular Histogram. It is significantly smaller than a regular Histogram when a large number of buckets are zero, which is a frequent occurence. It stores an individual vector for each field of non-zero buckets. Assuming index[0] = n, (index[0], count[0]) corresponds to the nth bucket.
Enums§
- Errors returned for histogram construction and operations.