QPR ProcessAnalyzer Expression Query Examples

This page contains example KPI Analyses.

Single ValueType Examples

Events grouped by event attributes

The following analysis calculates:

1. Number of cases
2. List of case names
3. List of distinct case names (single case appears only once in the list)

for each event attribute Color and Category.

{
  "Root": "Events",
  "Dimensions": [
    {
      "Name": "Color",
      "Expression": "Color"
    },
    {
      "Name": "Category",
      "Expression": "Category"
    }
  ],
  "Values": [
    {
      "Name": "Count",
      "Expression": "Count(_)" 
    },
    {
      "Name": "Cases",
      "Expression": "StringJoin(\",\", _.Case.Name)" 
    },
    {
      "Name": "DistinctCases",
      "Expression": "StringJoin(\",\", Distinct(_.Case.Name))" 
    }
  ],
  "Ordering": [ 
    {
      "Name": "Color",
      "Direction": "Ascending"
    },
    {
      "Name": "Category",
      "Direction": "Descending"
    }
  ]
}

Cases grouped by case attributes

Same analysis for for case attributes:

{
  "Root": "Cases",
  "Dimensions": [
    {
      "Name": "Color",
      "Expression": "Color"
    },
    {
      "Name": "Category",
      "Expression": "Category"
    }
  ],
  "Values": [
    {
      "Name": "Count",
      "Expression": "Count(_)" 
    },
    {
      "Name": "Cases",
      "Expression": "StringJoin(\",\", _.Name)" 
    }
    ],
  "Ordering": [ 
    {
      "Name": "Color",
      "Direction": "Ascending"
    },
    {
      "Name": "Category",
      "Direction": "Ascending"
    }
  ]
}

Return only dimension values

The following analysis returns a list of all dimensions but doesn't calculate any KPI's:

{
  "Root": "Cases",
  "Dimensions": [
    {
      "Name": "Region",
      "Expression": "Region"
    }
  ]
}

Return only single KPI value without dimensioning

The following analysis calculates a KPI for all filtered data but doesn't slice it to any dimensions:

{
  "Root": "Cases",
  "Dimensions": [],
  "Values": [
    {
      "Name": "Case count",
      "Expression": "Count(_)" 
    }
  ]
}

3 Grams

{
  "Root": "Cases",
  "Dimensions": [
    {
      "Name": "CaseId",
      "Expression": "Name"
    }
  ],
  "Values": [
    {
      "Expression": "
        Flatten(
          _
          .Array(
            StringJoin(\"->\", 
            Array(\"0\", 
              GetAt(0, Events).Type.Name, 
              If(
                Count(Events) > 1, 
                GetAt(1, Events).Type.Name, 
                \"0\"
                )
              )
            ),
            StringJoin(\"->\", 
              Events.Where(!IsNull(NextInCase))
              .Array(Type.Name, 
                NextInCase.Type.Name, 
                If(
                  !IsNull(NextInCase.NextInCase), 
                  NextInCase.NextInCase.Type.Name, 
                  \"0\"
                )
              )
            )
          )
        ).Where(_ != \"\")
      ",
      "DimensionOrderExpression": "OrderByValue(_)",
      "Type": "Pivot"
    }
  ],
  "Ordering": [ 
    {
      "Name": "CaseId",
      "Direction": "Ascending"
    }
  ]
}

Case duration distribution by hours

Following analysis shows case duration distribution by hours (works like the Duration Analysis)

{
  "Root": "Cases",
    "Dimensions": [
      {
        "Name": "Case Duration in Hours",
        "Expression": "Duration.TotalHours.Round(0)"
      }
    ],
    "Values": [
      {
        "Name": "Case Count",
        "Expression": "Count(_)" 
      }
    ],
    "Ordering": [ 
      {
        "Name": "Case Duration in Hours",
        "Direction": "Ascending"
    }
  ]
}

Monthly average case duration

The following analysis calculates monthly average case duration. Case start month needs to be a dimension and KPI is the average case duration (also rounding is used).

{
  "Root": "Cases",
  "Dimensions": [
    {
      "Name": "Start Month",
      "Expression": "StartTime.Month"
    }
  ],
  "Values": [
    {
      "Name": "Average Case Duration in Days",
      "Expression": "Round(Average(_.Duration.TotalSeconds) / 3600, 0)"
    }
  ],
  "Ordering": [
    {
      "Name": "Start Month",
      "Direction": "Ascending"
    }
  ]
}

Dynamic ValueType Examples

Counts of different occurred events by time

The following analysis shows how many of different kinds of events occurred by time (works like the Event Type Trend Analysis).

{
  "Root": "EventTypes",
  "Dimensions": [
    {
      "Name": "Event Name",
      "Expression": "Name"
    }
  ],
  "Values": [
    {
      "Name": "Event Count",
      "Expression": "GetAt(0, _.Count)" 
    },
    {
      "NameExpression": "",
      "ValueDimensionExpression": "TimeRange(DateTime(2012, 1, 1, 12), DateTime(2012, 1, 1, 13), TimeSpan(0, 0, 10))",
      "Expression": "GetAt(0, Count(_.Events.Where(TimeStamp.Round(TimeSpan(0, 0, 10)) == ValueDimension)))",
      "Type": "Dynamic"
    }
  ],
  "Ordering": [ 
    {
      "Name": "Event Name",
      "Direction": "Ascending"
    }
  ]
}

Pivot ValueType Examples

Event type ordering

{
  "Root": "Cases",
  "Dimensions": [
    {
      "Name": "CaseId",
      "Expression": "Name"
    }
  ],
  "Values": [
    {
      "Expression": "
        Flatten(
          _
          .Events
          .For(\"i\", IndexInCase + 1, i < Count(Case.Events), i + 1, 
            StringJoin(\">\", 
              Array(
                Type.Name, 
                GetAt(i, Case.Events).Type.Name
              )
            )
          )
        )
      ",
      "DimensionOrderExpression": "OrderByValue(_)",
      "Type": "Pivot"
    }
  ],
  "Ordering": [ 
    {
      "Name": "CaseId",
      "Direction": "Ascending"
    }
  ]
}

Repeats

{
  "Root": "Cases",
  "Dimensions": [
    {
      "Name": "CaseId",
      "Expression": "Name"
    }
  ],
  "Values": [
    {
      "Expression": "
        Flatten(
          _
          .FindRepeats(Events.Type.Name)
          .Repeat(
            Count(GetAt(1, _)) - 1, 
            StringJoin(\"->\", GetAt(0, _))
          )
        )
      ",
      "DimensionOrderExpression": "OrderByValue(_)",
      "Type": "Pivot"
    }
  ],
  "Ordering": [ 
    {
      "Name": "CaseId",
      "Direction": "Ascending"
    }
  ]
}

Workload per resouce

{
  "Root": "
TimeRange(DateTime(2012,1,1,12,0), DateTime(2012,1,1,13,0), TimeSpan(0, 0, 5))
", 
  "Dimensions": [
    { 
      "Name": "Time",
      "Expression": "_"
    }
  ], 
  "Values": [
    {
      "Type": "Pivot",
      "Expression": "
Def(\"ActiveEventUnitsAt\", \"at\", 
  Cases:GetAt(0, Events)
    .RecursiveFind(
      NextInCase,
      TimeStamp <= at 
      && (IsNull(NextInCase) || (NextInCase.TimeStamp > at))
    )
  .Unit
);
Let(\"CurrentTime\", 
  GetAt(
    0, 
    _
  ), 
  Flatten(
    EventLog
    .ActiveEventUnitsAt(
      CurrentTime
    )
  )
)
"
    }
  ]
}