时序数据降采样算法（LTOB, LTTB, LTD, OSI-PI Plot 等）的Java实现。
README - English
- LTOB: Largest-Triangel-One-Bucket
- LTTB: Largest-Triangel-Three-Bucket
- LTD: Largest-Triangel-Dynamic
- PIPLOT: OSI Soft PI 趋势图的降采样算法
- MAXMIN: 取最大、最小值的降采样算法
最大三角形系列算法的作者为Sveinn Steinarsson.
LTD算法的实现与原版有区别: 动态桶大小迭代的次数为 datasize / threshold / 10 且上限500.
使用 DSAlgorithms 类获得算法的实例。
List<Event> rawData = getData(); // Read data.
DownSamplingAlgorithm algorithm = DSAlgorithms.LTTB // get a algorithm.
List<Event> downsampled = algorithm.process(rawData, threshold); // do downsampling
MixedAlgorithm mixed = new MixedAlgorithm();
mixed.add(DSAlgorithms.LTTB, 1); // add LTTB with threshold * 1
mixed.add(new TimeGapAlgorithm(), 0.1); // add TG with threshold * 0.1
List<Event> downsampled = mixed.process(rawData, threshold); // do downsampling
运行test包下的 DownsampleTest类