[TimeSeries Explanation] Score is NaN

[lisahua]

System information

• OS version/distro: .Net 4.8
• .NET Version (eg., dotnet --info): ML.Net 1.5.0

Issue

• What did you do? Invoke var prediction = this.mlContext.AnomalyDetection.LocalizeRootCause(input);
• What happened? rootCauseItem.Score = double.NaN
• What did you expect? Should return a meaningful score (either 0 or 1)

Source code / logs

Please paste or attach the code or logs or traces that would be helpful to diagnose the issue you are reporting.

string jsonString = ""{\"AnomalyTimestamp\":\"1997-05-19T00:00:00\",\"AnomalyDimension\":{\"__Sandbox.Categories.CategoryName__\":\"\"},\"Slices\":[{\"TimeStamp\":\"1997-05-19T00:00:00\",\"Points\":[{\"Value\":8995.35,\"ExpectedValue\":-120.28187410606782,\"IsAnomaly\":true,\"Dimension\":{\"__Sandbox.Categories.CategoryName__\":\"Beverages\"},\"Delta\":9115.6318741060677},{\"Value\":1282.02,\"ExpectedValue\":1374.808978001977,\"IsAnomaly\":false,\"Dimension\":{\"__Sandbox.Categories.CategoryName__\":\"Condiments\"},\"Delta\":-92.788978001977057},{\"Value\":1686.7,\"ExpectedValue\":2014.8946291211198,\"IsAnomaly\":false,\"Dimension\":{\"__Sandbox.Categories.CategoryName__\":\"Confections\"},\"Delta\":-328.19462912111976},{\"Value\":174.15,\"ExpectedValue\":215.02809847811903,\"IsAnomaly\":false,\"Dimension\":{\"__Sandbox.Categories.CategoryName__\":\"Dairy Products\"},\"Delta\":-40.878098478119028},{\"Value\":12138.22,\"ExpectedValue\":9066.203,\"IsAnomaly\":true,\"Dimension\":{\"__Sandbox.Categories.CategoryName__\":\"\"},\"Delta\
":3072.017}]}],\"AggregateType\":0,\"AggregateSymbol\":\"\"}"
";
var input = JsonConvert.DeserializeObject<RootCauseLocalizationInput>(jsonString); // using Newtonsoft.Json
var prediction = this.mlContext.AnomalyDetection.LocalizeRootCause(input);

The input is like:

{
	"AnomalyTimestamp": "1997-05-19T00:00:00",
	"AnomalyDimension": {
		"__Sandbox.Categories.CategoryName__": ""
	},
	"Slices": [
		{
			"TimeStamp": "1997-05-19T00:00:00",
			"Points": [
				{
					"Value": 8995.35,
					"ExpectedValue": -120.28187410606782,
					"IsAnomaly": true,
					"Dimension": {
						"__Sandbox.Categories.CategoryName__": "Beverages"
					},
					"Delta": 9115.631874106068
				},
				{
					"Value": 1282.02,
					"ExpectedValue": 1374.808978001977,
					"IsAnomaly": false,
					"Dimension": {
						"__Sandbox.Categories.CategoryName__": "Condiments"
					},
					"Delta": -92.78897800197706
				},
				{
					"Value": 1686.7,
					"ExpectedValue": 2014.8946291211199,
					"IsAnomaly": false,
					"Dimension": {
						"__Sandbox.Categories.CategoryName__": "Confections"
					},
					"Delta": -328.19462912111978
				},
				{
					"Value": 174.15,
					"ExpectedValue": 215.02809847811904,
					"IsAnomaly": false,
					"Dimension": {
						"__Sandbox.Categories.CategoryName__": "Dairy Products"
					},
					"Delta": -40.87809847811903
				},
				{
					"Value": 12138.22,
					"ExpectedValue": 9066.203,
					"IsAnomaly": true,
					"Dimension": {
						"__Sandbox.Categories.CategoryName__": ""
					},
					"Delta": 3072.017
				}
			]
		}
	],
	"AggregateType": 0,
	"AggregateSymbol": ""
}

And the output:

{
	"Items": [
		{
			"Score": "NaN",
			"Path": [
				"__Sandbox.Categories.CategoryName__"
			],
			"Dimension": {
				"__Sandbox.Categories.CategoryName__": "Beverages"
			},
			"Direction": 0
		}
	]
}

[lisahua]

Fixed already in #5218