| /* |
| * |
| * Copyright 2014, Google Inc. |
| * All rights reserved. |
| * |
| * Redistribution and use in source and binary forms, with or without |
| * modification, are permitted provided that the following conditions are |
| * met: |
| * |
| * * Redistributions of source code must retain the above copyright |
| * notice, this list of conditions and the following disclaimer. |
| * * Redistributions in binary form must reproduce the above |
| * copyright notice, this list of conditions and the following disclaimer |
| * in the documentation and/or other materials provided with the |
| * distribution. |
| * * Neither the name of Google Inc. nor the names of its |
| * contributors may be used to endorse or promote products derived from |
| * this software without specific prior written permission. |
| * |
| * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
| * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
| * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR |
| * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT |
| * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, |
| * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT |
| * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, |
| * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY |
| * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
| * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE |
| * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
| * |
| */ |
| |
| #include <grpc/support/histogram.h> |
| |
| #include <math.h> |
| #include <stddef.h> |
| #include <string.h> |
| |
| #include <grpc/support/alloc.h> |
| #include <grpc/support/port_platform.h> |
| #include <grpc/support/log.h> |
| #include <grpc/support/useful.h> |
| |
| /* Histograms are stored with exponentially increasing bucket sizes. |
| The first bucket is [0, m) where m = 1 + resolution |
| Bucket n (n>=1) contains [m**n, m**(n+1)) |
| There are sufficient buckets to reach max_bucket_start */ |
| |
| struct gpr_histogram { |
| /* Sum of all values seen so far */ |
| double sum; |
| /* Sum of squares of all values seen so far */ |
| double sum_of_squares; |
| /* number of values seen so far */ |
| double count; |
| /* m in the description */ |
| double multiplier; |
| double one_on_log_multiplier; |
| /* minimum value seen */ |
| double min_seen; |
| /* maximum value seen */ |
| double max_seen; |
| /* maximum representable value */ |
| double max_possible; |
| /* number of buckets */ |
| size_t num_buckets; |
| /* the buckets themselves */ |
| gpr_uint32 *buckets; |
| }; |
| |
| /* determine a bucket index given a value - does no bounds checking */ |
| static size_t bucket_for_unchecked(gpr_histogram *h, double x) { |
| return (size_t)(log(x) * h->one_on_log_multiplier); |
| } |
| |
| /* bounds checked version of the above */ |
| static size_t bucket_for(gpr_histogram *h, double x) { |
| size_t bucket = bucket_for_unchecked(h, GPR_CLAMP(x, 0, h->max_possible)); |
| GPR_ASSERT(bucket < h->num_buckets); |
| return bucket; |
| } |
| |
| /* at what value does a bucket start? */ |
| static double bucket_start(gpr_histogram *h, double x) { |
| return pow(h->multiplier, x); |
| } |
| |
| gpr_histogram *gpr_histogram_create(double resolution, |
| double max_bucket_start) { |
| gpr_histogram *h = gpr_malloc(sizeof(gpr_histogram)); |
| GPR_ASSERT(resolution > 0.0); |
| GPR_ASSERT(max_bucket_start > resolution); |
| h->sum = 0.0; |
| h->sum_of_squares = 0.0; |
| h->multiplier = 1.0 + resolution; |
| h->one_on_log_multiplier = 1.0 / log(1.0 + resolution); |
| h->max_possible = max_bucket_start; |
| h->count = 0.0; |
| h->min_seen = max_bucket_start; |
| h->max_seen = 0.0; |
| h->num_buckets = bucket_for_unchecked(h, max_bucket_start) + 1; |
| GPR_ASSERT(h->num_buckets > 1); |
| GPR_ASSERT(h->num_buckets < 100000000); |
| h->buckets = gpr_malloc(sizeof(gpr_uint32) * h->num_buckets); |
| memset(h->buckets, 0, sizeof(gpr_uint32) * h->num_buckets); |
| return h; |
| } |
| |
| void gpr_histogram_destroy(gpr_histogram *h) { |
| gpr_free(h->buckets); |
| gpr_free(h); |
| } |
| |
| void gpr_histogram_add(gpr_histogram *h, double x) { |
| h->sum += x; |
| h->sum_of_squares += x * x; |
| h->count++; |
| if (x < h->min_seen) { |
| h->min_seen = x; |
| } |
| if (x > h->max_seen) { |
| h->max_seen = x; |
| } |
| h->buckets[bucket_for(h, x)]++; |
| } |
| |
| int gpr_histogram_merge(gpr_histogram *dst, gpr_histogram *src) { |
| size_t i; |
| if ((dst->num_buckets != src->num_buckets) || |
| (dst->multiplier != src->multiplier)) { |
| /* Fail because these histograms don't match */ |
| return 0; |
| } |
| dst->sum += src->sum; |
| dst->sum_of_squares += src->sum_of_squares; |
| dst->count += src->count; |
| if (src->min_seen < dst->min_seen) { |
| dst->min_seen = src->min_seen; |
| } |
| if (src->max_seen > dst->max_seen) { |
| dst->max_seen = src->max_seen; |
| } |
| for (i = 0; i < dst->num_buckets; i++) { |
| dst->buckets[i] += src->buckets[i]; |
| } |
| return 1; |
| } |
| |
| static double threshold_for_count_below(gpr_histogram *h, double count_below) { |
| double count_so_far; |
| double lower_bound; |
| double upper_bound; |
| size_t lower_idx; |
| size_t upper_idx; |
| |
| GPR_ASSERT(h->count >= 1); |
| |
| if (count_below <= 0) { |
| return h->min_seen; |
| } |
| if (count_below >= h->count) { |
| return h->max_seen; |
| } |
| |
| /* find the lowest bucket that gets us above count_below */ |
| count_so_far = 0.0; |
| for (lower_idx = 0; lower_idx < h->num_buckets; lower_idx++) { |
| count_so_far += h->buckets[lower_idx]; |
| if (count_so_far >= count_below) { |
| break; |
| } |
| } |
| if (count_so_far == count_below) { |
| /* this bucket hits the threshold exactly... we should be midway through |
| any run of zero values following the bucket */ |
| for (upper_idx = lower_idx + 1; upper_idx < h->num_buckets; upper_idx++) { |
| if (h->buckets[upper_idx]) { |
| break; |
| } |
| } |
| return (bucket_start(h, lower_idx) + bucket_start(h, upper_idx)) / 2.0; |
| } else { |
| /* treat values as uniform throughout the bucket, and find where this value |
| should lie */ |
| lower_bound = bucket_start(h, lower_idx); |
| upper_bound = bucket_start(h, lower_idx + 1); |
| return GPR_CLAMP(upper_bound - (upper_bound - lower_bound) * |
| (count_so_far - count_below) / |
| h->buckets[lower_idx], |
| h->min_seen, h->max_seen); |
| } |
| } |
| |
| double gpr_histogram_percentile(gpr_histogram *h, double percentile) { |
| return threshold_for_count_below(h, h->count * percentile / 100.0); |
| } |
| |
| double gpr_histogram_mean(gpr_histogram *h) { |
| GPR_ASSERT(h->count); |
| return h->sum / h->count; |
| } |
| |
| double gpr_histogram_stddev(gpr_histogram *h) { |
| return sqrt(gpr_histogram_variance(h)); |
| } |
| |
| double gpr_histogram_variance(gpr_histogram *h) { |
| if (h->count == 0) return 0.0; |
| return (h->sum_of_squares * h->count - h->sum * h->sum) / |
| (h->count * h->count); |
| } |
| |
| double gpr_histogram_maximum(gpr_histogram *h) { return h->max_seen; } |
| |
| double gpr_histogram_minimum(gpr_histogram *h) { return h->min_seen; } |
| |
| double gpr_histogram_count(gpr_histogram *h) { return h->count; } |
| |
| double gpr_histogram_sum(gpr_histogram *h) { return h->sum; } |
| |
| double gpr_histogram_sum_of_squares(gpr_histogram *h) { |
| return h->sum_of_squares; |
| } |