Calculated measures are values derived from an equation that uses previously defined measures. These equations can include process measures, output measures, other calculated measures, supported functions, and constants.

When to use calculated measures

Calculated measures are commonly used for convenience and consistency. For example, if a process includes measures for gross container weight and tare weight, you can create a calculated measure to automatically determine the net product weight.

Calculated measures are ideal when the value should update automatically whenever the source data changes.

Add a calculated measure to a process

Calculated measures are always associated with a specific process. To add a calculated measure to a process, open the process template.

  1. In the Measures section, select Add Calculated Measure.
  2. Enter a name and an optional description.
  3. Enter a valid equation.
    • Equation variables must already exist as process, output, or calculated measures.
    • Variable names must use the following format, with an underscore at the beginning and end of the name, _VariableName1_.
    • Select Show Variables to view available measures.
  4. Select OK, then save your changes.

For more information on process and output measures, go to Continuous measures.

Valid equations

Real-Time SPC supports basic mathematical operators and functions. Equations must follow the standard order of operations (PEMDAS). Parentheses are required for functions, exponents, and grouped operations.

As you type an equation, Real-Time SPC suggests supported functions.
Syntax Use
`abs(number)` Returns the absolute value of a number
`add(number1, number2, ...)` Adds multiple numbers together
`divide(number1, number2, ...)` Divides numbers sequentially
`max(number1, number2, ...)` Returns the maximum value
`mean(number1, number2, ...)` Calculates the average
`min(number1, number2, ...)` Returns the minimum value
`multiply(number1, number2, ...)` Multiplies multiple numbers
`round(number1, decimalnumber)` Rounds the number1 to number2 decimal places
`sqrt(number)` Returns the square root
`std(number1, number2, ...)` Calculates the standard deviation
`subtract(number1, number2, ...)` Subtracts numbers sequentially
Use the following syntax examples to create your own equations.
Calculate net weight
Subtract tare weight from gross weight.
  • subtract(_Gross Container Weight_,_Tare Weight_)
  • _Gross Container Weight_ - _Tare Weight_
  • _Gross Container Weight_ - 1.5
Calculate area
Multiply length and width.
  • multiply(_Length_,_Width_)
  • _Length_ * _Width_
Calculate the mean of several measures
mean(_Measure1_,_Measure2_,_Measure3_,_Measure4_)
Calculate the standard deviation of several measures
std(_Measure1_,_Measure2_,_Measure3_,_Measure4_)
Note

Only valid measures associated with the same process as the calculated measure are available for use in equations.

Deleted equation variables

If the equation for a calculated measure contains a deleted measure, the equation is no longer valid. The equation remains; however, the deleted measure is represented by a system id value. For example, the following valid equation, _Gross Container Weight_ - _Tare Weight_, becomes invalid when _Tare Weight_ is deleted from the process.

The invalid equation contains a system ID value instead of a valid measure name (_Gross Container Weight_ - _12_).

Real-Time SPC retains all previous calculations. Subsequent data collections will use a null value until the equation is valid again.

Sampling considerations

Calculated measures are available only for Data Collection Form data collections. Calculations are performed row by row.
  • All measures used in a calculated measure must be in the same sampling plan as the calculated measure.
  • The subgroup size for the calculated measure does not have to match the subgroup sizes for the variables in the defining equation.
  • Calculations are based on the smallest subgroup size among the measures used.

Example

If Net Product Weight has a subgroup size of n = 3, but Tare Weight has a subgroup size of n = 1, only one observation is calculated. As a result, an Xbar-R chart cannot display the data because at least two observations are required to calculate the range.
For more information, go to Overview of data collection methods for sampling plans.

Edit observations

You cannot edit calculated measures directly. Instead, edit one of the measures used in the equation.

When you edit an observation for a measure that is used in the calculated measure equation, Real-Time SPC automatically updates the following items.
  • Calculated measure values
  • Plotted points of the control chart
  • Tests for special causes
  • Summary and performance statistics
Note

If the original calculated data point was flagged and had assignable causes and/or corrective actions, these will remain if the point still exceeds the limits. If the point no longer exceeds the limits, the flag is removed; however, the assignable causes and correctives remain.

For more information, go to Edit observations and descriptors.