diff options
Diffstat (limited to 'src/benchmark_group.rs')
| -rwxr-xr-x | src/benchmark_group.rs | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/src/benchmark_group.rs b/src/benchmark_group.rs index c59cf56..ecb8258 100755 --- a/src/benchmark_group.rs +++ b/src/benchmark_group.rs @@ -3,6 +3,7 @@ use crate::benchmark::PartialBenchmarkConfig; use crate::connection::OutgoingMessage; use crate::measurement::Measurement; use crate::report::BenchmarkId as InternalBenchmarkId; +use crate::report::Report; use crate::report::ReportContext; use crate::routine::{Function, Routine}; use crate::{Bencher, Criterion, DurationExt, Mode, PlotConfiguration, SamplingMode, Throughput}; @@ -198,7 +199,7 @@ impl<'a, M: Measurement> BenchmarkGroup<'a, M> { /// noise. /// /// This presents a trade-off. By setting the significance level closer to 0.0, you can increase - /// the statistical robustness against noise, but it also weaken's Criterion.rs' ability to + /// the statistical robustness against noise, but it also weakens Criterion.rs' ability to /// detect small but real changes in the performance. By setting the significance level /// closer to 1.0, Criterion.rs will be more able to detect small true changes, but will also /// report more spurious differences. |