One of the properties that Codahale "Metrics" got is "biased."Can someone explain what this means?
public Histogram newHistogram(Class<?> klass, String name, boolean biased) Creates a new Histogram and registers it under the given class and name. Parameters: klass - the class which owns the metric name - the name of the metric biased - whether or not the histogram should be biased Returns: a new Histogram
If in doubt, look at the source:
/** * Uses an exponentially decaying sample of 1028 elements, which offers a 99.9% confidence * level with a 5% margin of error assuming a normal distribution, and an alpha factor of * 0.015, which heavily biases the sample to the past 5 minutes of measurements. */ BIASED { @Override public Sample newSample() { return new ExponentiallyDecayingSample(DEFAULT_SAMPLE_SIZE, DEFAULT_ALPHA); } };
A biased histogram is one that gives more weight to later measurements.