Acumen BI Control-Chart (I-Chart)

Provided by
...

Acumen BI Control-Chart (I-Chart)



Pricing and Setup

Provides a customisable, supported and easy to use I Chart extension for Qlik Sense. All you need to do is set the Dimension and Measure and let the extension handle the rest.

 

Table of Contents

Acumen BI - Control Chart (I-Chart) 1

Images. 2

I-Chart Configuration and Set Up. 2

Define the Dimension. 2

Define the Measure. 3

Define the I Chart Method. 4

Define the Measure Variables. 6

Define Styling for Other Measures. 8

Define Sorting. 9

Define Chart Appearance. 9

Definitions for Additional Measures Displayed in the Chart 10

Frequently Asked Questions. 14

Support 15

Roadmap, Feedback and Feature Requests. 16

Version Control 16

Terms and Conditions. 16

 

 

Acumen BI - Control Chart (I-Chart)

The Acumen BI Control Chart (I-Chart) provides a customisable, supported and easy to use I Chart extension for Qlik Sense. All you need to do is set the Dimension and Measure and let the extension handle the rest.

Key features include:

-          Outliers Highlighted

-          Runs 'above' and 'below' the average highlighted

-          Trends up and down highlighted

-          Ability to set break point to recalculate the average and upper and lower limits:

o   Manually

o   Automatically based on Runs and/ or Trend triggers

o   Not at all

Key configuration options include:

-          Customise Labels for Dimensions, Measures and Rules

-          Customise Line and point style for Dimensions, Measures and Rules

-          Pick the number of points to set the highlighting criteria for both Runs and Trends

Why the Acumen I-Chart

Control Charts are an analytical technique that plots data and process variation over time. It can:

-  Enable you to understand whether changes you are making are resulting in improvement.
-  Signal when a process may be going out of control.
-  Show where to look for sources of special cause variation.
-  Help us understand variation and in so doing guide us to take the most appropriate action.
-  Save time spent on unecessary 'corrective' action when a data point may just be a result of natural variation.

The Acumen I- Chart is a time series line chart with three reference lines and three rules that help you appreciate and highlight variation in the data.
-  The reference lines are:

o  Centre reference line: the average line.
o  Upper and lower reference lines: the process limits, also known as control limits. You don't set the limits as these are defined by the variability of data for the measure you are plotting. You can expect approximately 99% of data points to fall within the process limits.
-  The Three rules are:
o  Rule One: Outliers - Points that fall outside of the control limits.
o  Rule Two: Runs - Consecutive points above or below the average line. A run of values above or below the average (mean) line represents a trend that should not result from natural variation in the system.
o  Rule Three: Trend - Consecutive points increasing or decreasing. A run of six or more values showing continuous increase or decrease is a sign that something unusual is happening in the system.

-          Outliers Highlighted

 

I-Chart Configuration and Set Up

Define the Dimension

The dimension should be defined using the field that contains the 'date' values to be displayed on the x-axis of the I-Chart.

  1. Field: User Defined as per standard Qlik Field Definitions.
  2. Label: User Defined.
  3. Accepts One dimension only.

Define the Measure

The measure should be defined as the 'Events' that you are wanting to display on your control chart. Examples may include: Average Waiting Time, Total Attendances, Total Defects. This will be displayed on the y-axis of the control chart.

  1. Expression: User defined as per standard Qlik Measure Definitions.
  2. Label: User Defined.
  3. Number Formatting: User Defined.
  4. Show Data Points: yes/no. Displays or removes dots for each data point.
  5. Line Colour: Colour the line item using a simple colour picker or more advanced options.
  6. Line Style: Pick from Solid, Dashed or Dotted.
  7. Line Thickness: Adjusts line thickness.
  8. Accepts One measure only.

 

 

Define the I Chart Method

Under 'Measure Variables' you can define the method for the data points that are used in the Average and Upper and Lower Limit Calculations. The two options are:

1.       All Points Method

1.       This includes all Data Points in the Average and Moving Range Calculations up to a recalculation event.

2.       You then need to define the Minimum Points required before a Recalculation can automatically be applied. This is the minimum number of points required before the chart will automatically recalculate the Average and Moving Range, assuming you have chosen to automatically recalculate the Average and Moving range.

2.       Baseline Points Method

1.       Includes only an initial X baseline Data Points in the Average and Moving Range Calculations.

2.       You then need to define the number of baseline points you wish to use in the Average and Moving Range Calculations. This number of points will also be used again after a recalculation event.

1.       Calculation set to 'Auto'

Select a value from the Drop Down lists. These lists display all values from the Dimension

*Please see "What is Included in the Calculations.xlsx" for a detailed explanation of what data points are included in the calculations for each of the two methods '.

Define the Measure Variables

Under 'Measure Variables' you can configure:

2.       Rule 2 Runs: Sets the points above or below the average you are looking for in a run. You can choose between 5 and 9.

a)       Drives the data points highlighted as part of Rule 2 (Run).

b)      It also defines the Run points for the automatic 'Recalculation Event' scenarios, see point 3 below.

c)       Note: points must be above or below the average. A point equal to the average will break the run.

3.       Rule 3 Trends: Sets the consecutive points increasing or decreasing you are looking for in a trend. Choose between 5 and 9.

a)       Drives the data points highlighted as part of Rule 3 (Trend).

b)      It also defines the Trend points for the automatic 'Recalculation Event' scenarios, see point 3 below.

c)       Note: if the current point and the previous point are equal then this does not 'add to' or 'break' the trend and it carries on to check the next point.

4.       Define the 'Recalculation Event Rule: This determines if there is a recalculation break in the I Chart. Options include:

a)       Run or Trend: The I Chart will automatically insert a break if a Run or Trend is detected. Note: the Run will not display on the chart as the break is inserted at the start of the Run and the average is recalculated to take into account the data points that triggered the Run.

b)      Run Only: The I Chart will automatically insert a break if a Run is detected. Note: the Run will not display on the chart as the break is inserted at the start of the Run and the average is recalculated to take into account the data points that triggered the Run.

c)       Trend Only: The I Chart will automatically insert a break if a Trend is detected.

d)      No Recalculation Event: The I Chart will NOT automatically insert a break.

e)      Manual Break: Up to two manual breaks can be inserted when in 'Edit' mode for the sheet. For each Manual break you can input the break point by:

1.       Calculation set to 'Auto'

1.       Select a value from the Drop Down lists. These lists display all values from the Dimension. Note: you can select 'None' to only have one input value.

 

2.       Calculation set to ''Custom'

1.       You then have full flexibility to add in a calculation. Note: The calculation needs to result in ONE value that matches a SINGLE value in the dimension field. The match needs to include matching underlying date formatting.

2.       Example: = if(today() = '14/10/2018', '01/06/2010', '01/10/2010')

 

 

*Please see "What is Included in the Calculations.xlsx" for a detailed explanation of what data points are checked and where breaks are inserted for Run and Trend events.

Define Styling for Other Measures

Under 'Other Measures Formatting' you can configure the Styling for a number of other Measures and Rules that are automatically calculated once a Dimension and Measure have been defined. This includes the Average, Upper Control Limits, Lower Control Limits, three rules (Outliers, Run, Trend) and you can also add a Target line. Note:

  • Definitions for these are outlined in the section below 'Definitions for Additional Measures Displayed in the Chart'.

  • There are a number of variables that can be configured for the rules. This is outlined in the section above 'Define the Measure Variables'.

The Styling Options for the Average, Upper Control Limits, Lower Control Limits and Target are:

  1. Label: User Defined.
  2. Line Colour: Colour the line item using a simple colour picker or more advanced options.
  3. Line Style: Pick from Solid, Dashed or Dotted.
  4. Line Thickness: Adjusts line thickness.

 

The Styling Options for the three rules (Outliers, Run, Trend) are:

  1. Show Marker: Determines if the rule will be shown on the chart or not.
  2. Label: User Defined.
  3. Marker Colour: Colour the Marker using a simple colour picker or more advanced options.
  4. Marker Size: Adjusts the size of the marker.

 

Define Sorting

Chart must be sorted by the 'Dimension' (x-axis value) and in ascending value from lowest date on the left and highest date on the right of the chart.

 

Define Chart Appearance

Under 'Appearance' you can configure the appearance of the chart. This includes:

  1. General: Chart Title, Subtitle and Footnote. We recommend using this to communicate to users key points on how your I-Chart has been configured.
  2. Legend: Choose to display the legend or not and if so in what position.
  3. X-axis: Determine whether to show the title or not.
  4. y-axis: Determine whether to show the title or not.
  5. Calculation Condition: This allows you to determine if the chart should display or not. For example you may wish to set the minimum number of points in the Dimension for the chart to calculate. Example: =GetPossibleCount(Date) > 20

 

Definitions for Additional Measures Displayed in the Chart

Below we outline definitions for measures that are actually displayed in the chart over and above the Dimension and Measure outlined above. We have split this into the 'All Points' Method and the 'Baseline Points' Method.

 

I Chart Method: All Points Method

This includes all Data Points in the Average and Moving Range Calculations up to a recalculation event.

 

Default Label

Measure Summary / Definition

Average

Displays the average of the 'Events' measure. This average includes all values in the 'Events' measure up until any break point.

 

Lower Control Limit

Displays the Lower Control Limit calculated as: Average minus (2.66 *Moving Range).

 

The Average and Moving Range includes all values in the 'Events' measure up until any break point. Note: points outside the Upper Limit of the Initial Moving Range are excluded from the Moving Range calculation. See the section below 'How is the Moving Range calculated' for more information.

 

Upper Control Limit

This is the same definition as the Lower Control Limit apart from it is in ADDITION to the average not MINUS.

 

Rule 1 - A single point outside the control limits

An outlier is defined as a data point in the 'Events' measure outlined above being GREATER THAN the 'Upper Control Limit' value for any given Row

 

OR

 

a data point in the 'Events' measure outlined above being LESS THAN the 'Lower Control Limit' value for any given Row

 

 

Rule 2 - A run of X or more points in a row above or below the average

A Run is defined as:

 

X number of consecutive data points in the 'Events' measure being GREATER THAN (Above) the 'Average' Measure

 

Or

 

X number of consecutive data points in the 'Events' measure being LESS THAN (Below) the 'Average' Measure.

 

Where 'X' points is defined by the Qlik Application Developer. The default is set to 7.

 

Note: For the 'All Points Method' the calculation is taking the 'cumulative average' for the current row and then checking the next X points against this 'cumulative average' to see if they are above or below the 'current cumulative average'.

 

*Please see "What is Included in the Calculations.xlsx" for a detailed explanation of what data points are checked and where breaks are inserted for Run and Trend events.

 

Rule 3 - X consecutive points increasing (trend up) or decreasing (trend down)

A Trend is defined as:

 

X number of consecutive data points, where the data point in the 'Events' measure is GREATER THAN the previous data point in the 'Events' measure.

 

Or

 

X number of consecutive data points, where the data point in the 'Events' measure is LESS THAN the previous data point in the 'Events' measure.

 

Where 'X' points is defined by the Qlik Application Developer. The default is set to 6.

 

Note: If the current data point in the 'Events' measure and the previous data point in the 'Events' measure are EQUAL then this does not add to or end the Trend.

 

*Please see "What is Included in the Calculations.xlsx" for a detailed explanation of what data points are checked and where breaks are inserted for Run and Trend events.

 

 

 

I Chart Method: Baseline Points Method

Includes only an initial X baseline Data Points in the Average and Moving Range Calculations.

 

Default Label

Measure Summary / Definition

Average

Displays the average of the 'Events' measure. This average only includes an initial X baseline Data Points from the 'Events' measure. If there is a breakpoint then the new average only includes an initial X baseline Data Points from the 'Events' measure.

 

Lower Control Limit

Displays the Lower Control Limit calculated as: Average minus (2.66 *Moving Range).

 

The Average and Moving Range only includes an initial X baseline Data Points from the 'Events' measure. If there is a breakpoint then the new Average and Moving Range only includes an initial X baseline Data Points from the 'Events' measure.

Note: points outside the Upper Limit of the Initial Moving Range are excluded from the Moving Range calculation. See the section below 'How is the Moving Range calculated' for more information.

 

Upper Control Limit

This is the same definition as the Lower Control Limit apart from it is in ADDITION to the average not MINUS.

 

Rule 1 - A single point outside the control limits

An outlier is defined as a data point in the 'Events' measure outlined above being GREATER THAN the 'Upper Control Limit' value for any given Row

 

OR

 

a data point in the 'Events' measure outlined above being LESS THAN the 'Lower Control Limit' value for any given Row

 

 

Rule 2 - A run of X or more points in a row above or below the average

A Run is defined as:

 

X number of consecutive data points in the 'Events' measure being GREATER THAN (Above) the 'Average' Measure

 

Or

 

X number of consecutive data points in the 'Events' measure being LESS THAN (Below) the 'Average' Measure.

 

Where 'X' points is defined by the Qlik Application Developer. The default is set to 7.

 

*Please see "What is Included in the Calculations.xlsx" for a detailed explanation of what data points are checked and where breaks are inserted for Run and Trend events.

 

Rule 3 - X consecutive points increasing (trend up) or decreasing (trend down)

A Trend is defined as:

 

X number of consecutive data points, where the data point in the 'Events' measure is GREATER THAN the previous data point in the 'Events' measure.

 

Or

 

X number of consecutive data points, where the data point in the 'Events' measure is LESS THAN the previous data point in the 'Events' measure.

 

Where 'X' points is defined by the Qlik Application Developer. The default is set to 6.

 

Note: If the current data point in the 'Events' measure and the previous data point in the 'Events' measure are EQUAL then this does not add to or end the Trend.

 

*Please see "What is Included in the Calculations.xlsx" for a detailed explanation of what data points are checked and where breaks are inserted for Run and Trend events.

 

 

 

How is the Moving Range Calculated

The Moving Range measure mentioned in the two tables above is calculated based on the following steps:

1.       Calculate the 'Upper Limit of the Initial Moving Range' = 'Moving Range Average' * 3.27

2.       Exclude any data points that fall above the 'Upper Limit of the Initial Moving Range'

3.       Calculate the 'Moving Range' = Average of the Moving Range for points that have not been excluded.

Note: The above respects the selected 'I Chart Method' and any breaks (automatic or manual) in the chart. For example if the 'I Chart Method' is set at '15 Baseline Points' then only the first 15 points (or the first 15 after a break) are included in the Moving Range calculation outlined above.

 

Note:

  1. In regards to Rule 2: If a 'Event' data point is exactly equal to the 'Average' data point this will end the run. It should be noted this will be checking at a finer decimal point than what is displayed in the chart.
  2. In Regards to Rule 3: if the current point and the previous point are equal then this does not 'add to' or 'break' the trend and it carries on to check the next point.
  3. With the documentation provided Acumen BI have tried to be as open and transparent as possible to enable users to understand how measures are calculated and understand what data points are used in different methods. To date this is based on conversations with clients and potential clients. We appreciate there may be different takes on these methods and calculations and we welcome feedback on how this Extension can be improved. Please contact extensions@acumenbi.co.nz with your feedback.

 

Frequently Asked Questions

1.       Can we exclude certain points from the calculations?

a.       At this stage use native Qlik selections to select the points you want to exclude, you can then use native Qlik functionality to 'select excluded' which will in turn actually include the points for analysis. You could also exclude certain days or dates as part of the set analysis calculation for the measure.

 

2.       Can we exclude weekends?

a.       Use native Qlik functionality to add a Filter list showing 'Day of the Week'. This will allow users to select the days they wish to include in the charts analysis. You could also exclude certain days or dates as part of the set analysis calculation for the measure.

 

3.       Can we have different Manual breakpoints for different departments/business units.

a.       Use the Manual Break point method with a Custom Input on the value.

 

4.       How will users know how the I Chart has been configured?

a.       Use the Measure Labels and Chart Title, Subtitles and Footnotes to explain how each chart has been configured.

 

5.       How do I know what has triggered an automatic break?

a.       If the Recalculation Event is set to Trend Only then a Trend has triggered it.

b.       If the Recalculation Event is set to Run Only then a Run has triggered it.

c.       If the Recalculation Event is set to Run OR trend then you can tell this by:

                                                               i.      If the Trend rule is highlighted and the break is towards the end of this trend then a Trend is likely to have triggered the recalculation.

                                                             ii.      If there is no highlighting - then it will likely be a Run that has triggered the recalculation. The Run will not be highlighted as the break is put at the start of the run so once it is identified and recalculated it automatically clears the Run Highlighting as it is no longer a run due to the new Avg.

 

6.       Why is there Trend or Run and it has not automatically recalculated the average and upper and lower limits.

a.       The chart may be set to No Recalculation Event or Manual Input

b.       There will also be some cases where you may see a Run or Trend and it does not automatically recalculate. This is because there is minimum number of points that are required before the chart will automatically recalculate.

 

7.       Why do numbers not match the NHS statistical process control tool?

a.       The NHS Statistical Process Control tool can be found here: https://improvement.nhs.uk/resources/statistical-process-control-tool/

b.       The Reason the Upper Control Limit and Lower Control Limit Values may not match are the Acumen I Chart ignores 'high values' from the Moving Range calculation. From what we can see the 'NHS Statistical Process Control tool' flags these values however includes them in the Moving Range calculation. This can be changed on the Acumen I Chart however we would like feedback from the community as to the preferred method of calculation here. Please contact us at extensions@acumenbi.co.nz with your feedback.

 

8.       Why do numbers not match another Process Control Tool?

a.       In the first instance check if the break is set at the start or end of a Run.

b.       Please contact us at extensions@acumenbi.co.nz with your feedback and will investigate this.

 

9.       Can the user add in manual breaks?

a.       This could be achieved by using a �Custom� manual break calculation and using a variable in the calculation. You would then need to provide a method for the user to set the value in the variable.

Support

-          Product currently in BETA

-          Currently Out of Scope

o   Max 100 values on the X Axis

o   Support for Master Dimension/Measure

o   Support for Master Visualization

o   Support for PDF/IMG/XLSX export

o   Support for Stories

o   nPrinting, all versions.

o   Qlik Analytics Platform (QAP). Note: we have not tested this so it may work.

o   Versions prior to June 2018. Basic testing has shown it is generally compatible however we are still to define if early Qlik Sense versions will be supported officially.

-          Hours

o   Standard business hours for Auckland, New Zealand. Alternative arrangements can be discussed on request.

-          Contact us at extensions@acumenbi.co.nz with your support requests and bug reporting.

 

Roadmap, Feedback and Feature Requests

Contact us at extensions@acumenbi.co.nz with your feedback, feature requests, support requests and bug reporting. We also welcome suggestions on new charts and extensions you would like to see developed. This will feed into our wider development Road Map.

We understand there are further rules that could be included and that globally there are several different versions of the I-Chart and how the Averages and Rules might be calculated. - We welcome your feedback on what you would like to see developed. We will incorporate as much flexibility as possible and use feedback to develop our future Road Map.

Version Control

-          Product currently in BETA

Terms and Conditions

By using the Acumen BI Qlik Sense Extensions and Services you acknowledge and agree that all such use constitute's your acceptance of the following Terms & Conditions:

- Marketplace Terms & Conditions

- Acumen Extensions Terms & Conditions

 

...
Acumen BI Control-Chart (I-Chart)

60 Day Trial

Free to try it out
Free to Trial
Unlimited Users
Unlimited Projects
100% on-premise
Download for Free Marketplace Terms & Conditions