UTI: Add regressions and progressions

Change-Id: Ic5febe8fc1bd5ccd9699e73003783484240cbd07
Signed-off-by: Tibor Frank <tifrank@cisco.com>

3 files changed, 89 insertions, 5 deletions

diff --git a/resources/tools/dash/app/pal/data/data.py b/resources/tools/dash/app/pal/data/data.py
index f2c02acc63..296db024c0 100644
--- a/resources/tools/dash/app/pal/data/data.py
+++ b/resources/tools/dash/app/pal/data/data.py
@@ -213,15 +213,15 @@ class Data:
                 days=days
             ),
             self._create_dataframe_from_parquet(
-                path=self._get_path("statistics-trending"),
+                path=self._get_path("statistics-trending-mrr"),
                 partition_filter=l_mrr,
-                columns=self._get_columns("statistics-trending"),
+                columns=self._get_columns("statistics-trending-mrr"),
                 days=days
             ),
             self._create_dataframe_from_parquet(
-                path=self._get_path("statistics-trending"),
+                path=self._get_path("statistics-trending-ndrpdr"),
                 partition_filter=l_ndrpdr,
-                columns=self._get_columns("statistics-trending"),
+                columns=self._get_columns("statistics-trending-ndrpdr"),
                 days=days
             )
         )
diff --git a/resources/tools/dash/app/pal/data/data.yaml b/resources/tools/dash/app/pal/data/data.yaml
index 2585ef0e84..59533f97a4 100644
--- a/resources/tools/dash/app/pal/data/data.yaml
+++ b/resources/tools/dash/app/pal/data/data.yaml
@@ -5,7 +5,7 @@ statistics:
     - build
     - start_time
     - duration
-statistics-trending:
+statistics-trending-ndrpdr:
   path: s3://fdio-docs-s3-cloudfront-index/csit/parquet/trending
   columns:
     - job
@@ -13,8 +13,23 @@ statistics-trending:
     - dut_type
     - dut_version
     - hosts
+    - start_time
+    - passed
+    - test_id
+    - result_pdr_lower_rate_value
+    - result_ndr_lower_rate_value
+statistics-trending-mrr:
+  path: s3://fdio-docs-s3-cloudfront-index/csit/parquet/trending
+  columns:
+    - job
+    - build
+    - dut_type
+    - dut_version
+    - hosts
+    - start_time
     - passed
     - test_id
+    - result_receive_rate_rate_avg
 trending-mrr:
   path: s3://fdio-docs-s3-cloudfront-index/csit/parquet/trending
   columns:
diff --git a/resources/tools/dash/app/pal/data/utils.py b/resources/tools/dash/app/pal/data/utils.py
new file mode 100644
index 0000000000..63c9c1aaa4
--- /dev/null
+++ b/resources/tools/dash/app/pal/data/utils.py
@@ -0,0 +1,69 @@
+# Copyright (c) 2022 Cisco and/or its affiliates.
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at:
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""
+"""
+
+from numpy import isnan
+
+from ..jumpavg import classify
+
+
+def classify_anomalies(data):
+    """Process the data and return anomalies and trending values.
+
+    Gather data into groups with average as trend value.
+    Decorate values within groups to be normal,
+    the first value of changed average as a regression, or a progression.
+
+    :param data: Full data set with unavailable samples replaced by nan.
+    :type data: OrderedDict
+    :returns: Classification and trend values
+    :rtype: 3-tuple, list of strings, list of floats and list of floats
+    """
+    # NaN means something went wrong.
+    # Use 0.0 to cause that being reported as a severe regression.
+    bare_data = [0.0 if isnan(sample) else sample for sample in data.values()]
+    # TODO: Make BitCountingGroupList a subclass of list again?
+    group_list = classify(bare_data).group_list
+    group_list.reverse()  # Just to use .pop() for FIFO.
+    classification = list()
+    avgs = list()
+    stdevs = list()
+    active_group = None
+    values_left = 0
+    avg = 0.0
+    stdv = 0.0
+    for sample in data.values():
+        if isnan(sample):
+            classification.append("outlier")
+            avgs.append(sample)
+            stdevs.append(sample)
+            continue
+        if values_left < 1 or active_group is None:
+            values_left = 0
+            while values_left < 1:  # Ignore empty groups (should not happen).
+                active_group = group_list.pop()
+                values_left = len(active_group.run_list)
+            avg = active_group.stats.avg
+            stdv = active_group.stats.stdev
+            classification.append(active_group.comment)
+            avgs.append(avg)
+            stdevs.append(stdv)
+            values_left -= 1
+            continue
+        classification.append("normal")
+        avgs.append(avg)
+        stdevs.append(stdv)
+        values_left -= 1
+    return classification, avgs, stdevs