Fix prediction (#118)

analytics/analytic_unit_manager.py

@@ -30,14 +30,16 @@ def ensure_worker(analytic_unit_id, analytic_unit_type) -> AnalyticUnitWorker:
 async def handle_analytic_task(task):
     try:
         payload = task['payload']
-        payload['data'] = pd.DataFrame(payload['data'], columns = ['timestamp', 'value'])
+
         worker = ensure_worker(task['analyticUnitId'], payload['pattern'])
 
+        data = pd.DataFrame(payload['data'], columns=['timestamp', 'value'])
+        data['timestamp'] = pd.to_datetime(data['timestamp'])
         result_payload = {}
         if task['type'] == "LEARN":
-            await worker.do_learn(AnalyticUnitId, payload)
+            await worker.do_learn(payload['segments'], data)
         elif task['type'] == "PREDICT":
-            result_payload = await worker.do_predict(AnalyticUnitId, payload)
+            result_payload = await worker.do_predict(data)
         else:
             raise ValueError('Unknown task type "%s"' % task['type'])
         return {

analytics/analytic_unit_worker.py

@@ -1,6 +1,7 @@
 import config
 import detectors
 import logging
+import pandas as pd
 
 
 logger = logging.getLogger('AnalyticUnitWorker')
@@ -12,16 +13,10 @@ class AnalyticUnitWorker:
         self.analytic_unit_id = analytic_unit_id
         self.detector = detector
 
-    async def do_learn(self, analytic_unit_id, payload) -> None:
-        pattern = payload['pattern']
-        segments = payload['segments']
-        data = payload['data'] # [time, value][]
+    async def do_learn(self, segments: list, data: pd.DataFrame) -> None:
         await self.detector.train(data, segments)
 
-    async def do_predict(self, analytic_unit_id, payload):
-        pattern = payload['pattern']
-        data = payload['data'] # [time, value][]
-
+    async def do_predict(self, data: pd.DataFrame):
         segments, last_prediction_time = await self.detector.predict(data)
         return {
             'segments': segments,

analytics/detectors/pattern_detector.py

@@ -36,24 +36,17 @@ class PatternDetector(Detector):
         # TODO: save model after fit
         return 0
 
-    async def predict(self, data):
+    async def predict(self, dataframe: pd.DataFrame):
+        predicted_indexes = await self.model.predict(dataframe)
 
-        start_index = self.data_prov.get_upper_bound(last_prediction_time)
-        start_index = max(0, start_index - window_size)
-        dataframe = self.data_prov.get_data_range(start_index)
-
-        predicted_indexes = self.model.predict(dataframe)
-        predicted_indexes = [(x, y) for (x, y) in predicted_indexes if x >= start_index and y >= start_index]
-
-        predicted_times = self.data_prov.inverse_transform_indexes(predicted_indexes)
         segments = []
-        for time_value in predicted_times:
-            ts1 = int(time_value[0].timestamp() * 1000)
-            ts2 = int(time_value[1].timestamp() * 1000)
-            segments.append({
-                'start': min(ts1, ts2),
-                'finish': max(ts1, ts2)
-            })
+        # for time_value in predicted_times:
+        #     ts1 = int(time_value[0].timestamp() * 1000)
+        #     ts2 = int(time_value[1].timestamp() * 1000)
+        #     segments.append({
+        #         'start': min(ts1, ts2),
+        #         'finish': max(ts1, ts2)
+        #     })
 
         last_dataframe_time = dataframe.iloc[-1]['timestamp']
         last_prediction_time = int(last_dataframe_time.timestamp() * 1000)

analytics/models/step_model.py

@@ -6,7 +6,7 @@ from scipy.signal import argrelextrema
 
 import utils
 import numpy as np
-import pickle
+import pandas as pd
 
 
 class StepModel(Model):
@@ -30,8 +30,8 @@ class StepModel(Model):
         convolve_list = []
         for segment in segments:
             if segment['labeled']:
-                segment_from_index = utils.timestamp_to_index(dataframe, segment['from'])
-                segment_to_index = utils.timestamp_to_index(dataframe, segment['to'])
+                segment_from_index = utils.timestamp_to_index(dataframe, pd.to_datetime(segment['from']))
+                segment_to_index = utils.timestamp_to_index(dataframe, pd.to_datetime(segment['to']))
 
                 segment_data = data[segment_from_index : segment_to_index + 1]
                 segment_min = min(segment_data)

server/src/controllers/analytics_controller.ts

@@ -28,7 +28,7 @@ function onTaskResult(taskResult: TaskResult) {
       resolver(taskResult);
       taskResolvers.delete(id);
     } else {
-      throw new Error(`TaskResut [${id}] has no resolver`);
+      throw new Error(`TaskResult [${id}] has no resolver`);
     }
   }
 }
@@ -124,7 +124,6 @@ export async function runLearning(id: AnalyticUnit.AnalyticUnitId) {
       AnalyticUnit.setPredictionTime(id, lastPredictionTime)
     ]);
     await AnalyticUnit.setStatus(id, AnalyticUnit.AnalyticUnitStatus.READY);
-
   } catch (err) {
     let message = err.message || JSON.stringify(err);
     await AnalyticUnit.setStatus(id, AnalyticUnit.AnalyticUnitStatus.FAILED, message);
@@ -139,7 +138,7 @@ async function processLearningResult(taskResult: any): Promise<{
   lastPredictionTime: number,
   segments: Segment.Segment[]
 }> {
-  if(taskResult.status !== 'SUCCESS') {
+  if(taskResult.status !== AnalyticUnit.AnalyticUnitStatus.SUCCESS) {
     return Promise.reject(taskResult.error);
   }
   console.log(taskResult)
@@ -153,8 +152,8 @@ async function processLearningResult(taskResult: any): Promise<{
   // }
 
   return {
-    lastPredictionTime: +taskResult.lastPredictionTime,
-    segments: taskResult.segments.map(Segment.Segment.fromObject)
+    lastPredictionTime: 0,
+    segments: []
   };
 
 }
@@ -163,18 +162,27 @@ export async function runPredict(id: AnalyticUnit.AnalyticUnitId) {
   let unit = await AnalyticUnit.findById(id);
   let pattern = unit.type;
 
+  let segments = await Segment.findMany(id, { labeled: true });
+  if (segments.length < 2) {
+    throw new Error('Need at least 2 labeled segments');
+  }
+
+  let { from, to } = getQueryRangeForLearningBySegments(segments);
+  let data = await queryByMetric(unit.metric, unit.panelUrl, from, to);
+  if (data.length === 0) {
+    throw new Error('Empty data to predict on');
+  }
+
   let task = new AnalyticsTask(
     id,
     AnalyticsTaskType.PREDICT,
-    { pattern, lastPredictionTime: unit.lastPredictionTime }
+    { pattern, lastPredictionTime: unit.lastPredictionTime, data }
   );
   let result = await runTask(task);
-
-  if(result.status === 'FAILED') {
+  if(result.status === AnalyticUnit.AnalyticUnitStatus.FAILED) {
     return [];
   }
   // Merging segments
-  let segments = await Segment.findMany(id, { labeled: true });
   if(segments.length > 0 && result.segments.length > 0) {
     let lastOldSegment = segments[segments.length - 1];
     let firstNewSegment = result.segments[0];
@@ -214,6 +222,6 @@ export async function updateSegments(
   ]);
   // TODO: move setting status somehow "inside" learning
   await AnalyticUnit.setStatus(id, AnalyticUnit.AnalyticUnitStatus.PENDING);
-  runLearning(id);
+  runLearning(id).then(() => runPredict(id));
   return { addedIds, removed };
 }