From 86c4433b65f8753e7f71f8380c6593aab4be6415 Mon Sep 17 00:00:00 2001
From: Evgeny Smyshlyaev <amper43@gmail.com>
Date: Fri, 1 Mar 2019 14:02:16 +0300
Subject: [PATCH] Analytic bucket size #446 (#451)

---
 analytics/analytics/buckets/data_bucket.py        |  3 ++-
 analytics/analytics/detectors/pattern_detector.py | 12 ++++++------
 2 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/analytics/analytics/buckets/data_bucket.py b/analytics/analytics/buckets/data_bucket.py
index ee791d1..6606984 100644
--- a/analytics/analytics/buckets/data_bucket.py
+++ b/analytics/analytics/buckets/data_bucket.py
@@ -11,4 +11,5 @@ class DataBucket(object):
         self.data = self.data.append(data, ignore_index=True)
 
     def drop_data(self, count: int):
-        self.data = self.data.iloc[count:]
+        if count > 0:
+            self.data = self.data.iloc[count:]
diff --git a/analytics/analytics/detectors/pattern_detector.py b/analytics/analytics/detectors/pattern_detector.py
index 84f5290..bdbc560 100644
--- a/analytics/analytics/detectors/pattern_detector.py
+++ b/analytics/analytics/detectors/pattern_detector.py
@@ -37,8 +37,7 @@ class PatternDetector(Detector):
         self.analytic_unit_id = analytic_unit_id
         self.pattern_type = pattern_type
         self.model = resolve_model_by_pattern(self.pattern_type)
-        self.max_window_size = 150
-        self.window_size = 0
+        self.min_bucket_size = 150
         self.bucket = DataBucket()
 
     def train(self, dataframe: pd.DataFrame, segments: list, cache: Optional[models.ModelCache]) -> models.ModelCache:
@@ -71,12 +70,13 @@ class PatternDetector(Detector):
             return None
 
         self.bucket.receive_data(data_without_nan)
-        if cache and self.window_size == 0:
-            self.window_size = cache['WINDOW_SIZE']
+        if not cache: cache = {}
+        bucket_size = max(cache.get('WINDOW_SIZE', 0) * 3, self.min_bucket_size)
 
         res = self.detect(self.bucket.data, cache)
-        if len(self.bucket.data) >= self.window_size and cache != None:
-            excess_data = len(self.bucket.data) - self.max_window_size
+
+        if len(self.bucket.data) > bucket_size:
+            excess_data = len(self.bucket.data) - bucket_size
             self.bucket.drop_data(excess_data)
         
         if res: