WHAT ARE THE DIFFERENT OPTIONS IN SELECT FUNCTIONS ?

You have different functions to display in a time plot graph:

/!\ In a chart it is not possible to have SUM alone to add a marker, it is necessary to add the average
/!\ Beyond 50 000 points, it is not possible to make a graph with points connected in a curve, there are too many, and it would be totally unreadable

• Values : Displays the values of the variable
• Average: Displays the average value of the variable
  • Parameter: "By" (optional). In "By" you can select a Category variable and the average will be calculated for each period of time where this variable remains the same (for example : average by shifts if you have a variable giving shifts names).
• Moving Average: Displays the average of the values on a number of points before and after this point.
  Be careful : the parameters is the number of points. For example, if you have a database with 1 point = 1 minute, by default it is the moving average between 20 min before and 20 min after.
  • Parameters: Number of Points Before / Number of Points After
• Moving Range: Displays the difference between the maximal and minimal values of the points between a parametrized number of points before and after this point.
  • Parameters: Number of Points Before / Number of Points After
• Median: Displays the median value of the variable.
• Maximum: Displays the maximum value of the variable.
• Minimum: Displays the minimum value of the variable.
• Automatic Cusum: Displays the Cumulated Sum of the differences between the values of the points and the average.
  This allows to see the directions of deviations.
  For example: if your graph had (only !) 5 points with the following values in order : 9, 5, 2, 1, 3.
  The average of these is 4.
  So the Cusum will display the following values :
  5 (=9-4)
  6 (=5+5-4)
  4(=6+2-4)
  1 (=4+1-4)
  0(=1+3-4)
  • Parameter: "By" (optional). In "By", you can select a Category variable and the average will be calculated for each period of time where this variable is constant.
• Custom Cusum: Works the same way as the automatic cusum. Except that instead of the average, it takes the parameter in Reference to calculate the Cumulated Sum of the differences between the values of the points and the Reference.
  • Parameter: Reference
• Confidence Interval: Displays a colored horizontal range between (Average – Number of Sigma x Standard Deviation) and (Average + Number of Sigma x Standard Deviation).

If your variable has a normal distribution :

• 1 sigma : 68% of the values are in the Interval.
• 2 sigma : 95.5% of the values are in the Interval.
• 3 sigma : 99.7% of the values are in the Interval.
  • Parameter: Number of Sigmas
• Specification: Displays 2 red lines for the specified Lower and Upper limits and a black line with the average.
  • Parameters: LSL (Lower Specification Limit) and USL (Upper Specification Limit)
    • If no information is entered, the default value will be LSL = Mean-2STD ; USL = Mean+2STD
  • Range displayed (gray/shaded with default colors) is the Shewhart Natural Process limits as calculated by adding and subtracting (2.66 * Average Moving Range) from the Mean
• Coefficient of variation: Available in the Aggregate plot.
• Gaussian: Available in a Histogram plot.
• Affine regression: Available in a Multiplot.
• Linear regression: Available in a Multiplot.
• Exponential trend: Available in a Multiplot.
• Logarithmic trend: Available in a Multiplot.
• Power trend: Available in a Multiplot.