Add test data, update README

Author: rmoff

.gitignore

@@ -21,4 +21,5 @@ dist/
 .project
 .settings
 .tern-port
-ui/
+ui/
+logs/

EcdSensorData.avro

@@ -0,0 +1,29 @@
+{
+        "namespace": "ecd-sensor-data",
+        "name": "HealthSensorInputTopic",
+        "type": "record",
+        "fields": [
+        
+        
+                {"name": "eventid", "type": {
+                    "type": "long",
+                    "arg.properties": {
+                       "iteration": { "start": 1}
+                    }
+                }},
+                        
+                
+    {
+        "name": "heartbeat",
+        "type":
+          {
+            "type": "string",
+            "arg.properties": {
+              "regex": "([2-7]\\.[0-9][1-9]\\#){209}[2-7]\\.[0-9][1-9]"
+            }
+          }
+      }
+                          
+                
+    ]
+}

README.md

@@ -71,8 +71,7 @@ Send a sample message (returns a prediction of 1.2104138026620321):
 Create derived stream in KSQL:
 
   CREATE STREAM AnomalyDetection WITH (VALUE_FORMAT='AVRO') AS \
-  SELECT TIMESTAMPTOSTRING(ROWTIME, 'yyyy-MM-dd HH:mm:ss') AS EVENT_TS, \
-  eventid, \
+  SELECT eventid, sensorinput, \
   CAST (anomaly(sensorinput) AS DOUBLE) as Anomaly \
   FROM healthsensor;
 
@@ -125,7 +124,8 @@ Generates all readings with same/close timestamp though. To spread out over time
 
 Run continually:
 
-  cat ecg_discord_test.msgs | pv -q -L 1000| kafkacat -b localhost:9021 -P -t HealthSensorInputTopic
+  cd test-data
+  ./stream_loop_of_test_data_into_kafka.sh
 
 ### Generating random test data
 
@@ -135,6 +135,15 @@ Run continually:
 
 This uses the ksql-datagen tool (part of KSQL project) to generate test data. Whilst it provides random data, it's not very realistic to real-world data since it is truly random, rather than following a particular realistic pattern.
 
+### Change anomaly threshold
+
+  TERMINATE CSAS_ANOMALYDETECTIONBREACH;
+  DROP STREAM ANOMALYDETECTIONBREACH;
+  CREATE STREAM AnomalyDetectionBreach AS \
+  SELECT * FROM AnomalyDetectionAvro2 \
+    WHERE Anomaly >2;
+
+
 ## Stream to Elasticsearch
 
 Create a Kafka Connect sink to stream all scored events to Elasticsearch: