Home / Statistical Tools / Control Charts / T Chart / Options
Options¶
Want to build one first?
How-To works through a finished chart step by step, from sample data you copy off the page straight into Excel. The data has a point that is out of control, so the result has something to read rather than twenty points in a row.
How-To with GroupBy does the same thing with a GroupBy column, so one run produces a separate chart per group.
This page covers the T Chart dialog. It is not built from the same options kit as the variable charts, so several groups you may expect are absent rather than disabled, and each absence is called out below at the point where you would look for it. The pane that opens beside the finished chart has its own page, Task Pane. What each option means statistically is on the Math Details page.
The Dialog¶
The dialog is titled T Chart and has three tabs: Data, Limits & Splits, and Rules & Display. There is no Chart: variant row above them: one panel, no variants. The buttons along the bottom are Back, Help, Cancel and Finish.
There is no Capability tab, and that is not a hidden tab: no specification limit, no target, no capability index and no histogram can reach a T chart's output, in the engine as well as in the dialog.
Also absent: the Transformation group, the Control Limits limit-type radios and manual limits grid, the usual five-box Historical Values group, and the Estimate Sigma group. What replaces the last two is the distribution fitting described below.
Data Tab¶
The calculation type sits at the top as Excel and GroupBy, with a Data Types button beside it. Below that are two slot lists, Data Columns (required) and X-Axis Labels (Optional), and under GroupBy a region with Available Items and GroupBy Order.
Four things are specific to this chart:
- Four data types are accepted: Continuous, Integer, Count and DateTime. That is wider than the G chart, which takes only Count and DateTime, and it is the point of the T chart: the interval can be a measurement with decimals.
- Checking more than one data column produces more than one chart, one per column. There is no subgroup concept: one row is one event.
- The minimum is 3 intervals, or 4 date-times. Four stamps produce three intervals, because the first only establishes the starting point.
- A DateTime column gets a Time Unit drop-down of its own, beside the column in the Data Columns list. It appears only for DateTime columns.
Intervals or date-times¶
An interval column holds the elapsed times directly, in whatever unit you measured.
A DateTime column holds the moment each event occurred, and Quantum XL does the subtraction. Three details matter:
- The Time Unit drop-down converts the difference: Days, Hours, Minutes or Seconds, with Days the default. The factors are exactly what they look like: one day, a twenty-fourth of a day, a 1440th, an 86400th.
- The result keeps its fraction. Nothing is truncated, so two events 3.9 days apart contribute 3.9 days, or 93.6 hours. This is the difference that separates this chart from the G Chart, which floors the same difference to 3.
- The stamps must be in chronological order. One out-of-order row stops the analysis with a message naming both row numbers and both dates. And the first stamp produces no point, so N stamps give N minus 1 intervals.
The Y-axis title follows the unit you chose: Days Between Events, Hours Between Events and so on, or plain Time Between Events when the column was numeric.
Limits & Splits Tab¶
Three groups: Control Limits, Baseline Estimation and Split Control Limits.
Control Limits¶
| Control | Type | Default | What it does |
|---|---|---|---|
| Number of StDevs: | text box | 3 | Selects which quantiles of the fitted distribution the lines are drawn at |
| Fit exponential distribution only | check box | off | Fixes the shape at 1 and makes the scale the sample mean |
| Historical Weibull α (scale): | text box | blank | Supplies the scale outright |
| Historical Weibull β (shape): | text box | blank | Supplies the shape outright |
The sigma multiplier does not become a distance. There is no standard deviation being multiplied. The number is converted into tail probabilities through the normal distribution and the lines are the Weibull quantiles at those probabilities: at the default 3, the 0.135th and 99.865th percentiles. It must be greater than zero.
Fit exponential distribution only is the setting to understand first, because it changes what the chart is capable of telling you and it also changes how reproducible the numbers are:
- Cleared, the default, both Weibull parameters are estimated from your intervals by an iterative fit. A shape above 1 means the longer you wait the more likely the next event becomes; below 1, the opposite. That is real information a rare-event record can carry.
- Ticked, the shape is 1 and the scale is the plain average of your intervals. Every line on the chart is then a multiple of that average, and you can check the whole chart with a calculator.
Ticking the box clears and disables the Historical Weibull β box, since the shape is being forced to 1. The α box stays available, so supplying a scale while the shape is forced to 1 is a valid combination: the chart is then fully determined by one number you typed.
In general-Weibull mode the two historical boxes are all or nothing: supply both and neither is estimated, supply neither and both are estimated, supply exactly one and the analysis stops with an invalid-override error. Either value must be greater than zero.
Baseline Estimation¶
| Control | Type | Default | What it does |
|---|---|---|---|
| All Subgroups | radio button | selected | Fits from every event |
| First N Subgroups | radio button | not selected | Fits from the first N events only |
The count is a count of events, not worksheet rows, and only the fit uses it: every interval is still plotted. The minimum is 3, or 4 when the selected column is a DateTime column. The lines from those first N events extend across the whole chart.
Supplying both historical parameters makes this group irrelevant, because nothing is being fitted.
Split Control Limits¶
| Control | Type | Default | What it does |
|---|---|---|---|
| None | radio button | selected | One set of limits for the whole chart |
| Column | radio button | not selected | A new phase begins wherever the chosen column's value changes |
| Manual Breaks | radio button | not selected | A new phase begins at row numbers you list |
| Split Column: | drop-down | empty | The column read under Column mode |
| Manual Break Rows (1-based, start of split): | text box + Add / Remove Selected | empty | The list of phase starting rows |
| Split Labels (shown beside each divider): | grid | empty | One editable label per phase |
Each phase is fitted separately, so every phase gets its own shape, scale, median and limits. A divider is drawn at each boundary with the label you typed, and the summary table gains a Per-Phase Breakdown block. A phase needs enough events to fit on its own, which is the practical limit on how finely a rare-event record can be split.
Rules & Display Tab¶
Out-of-Control Rules¶
Only Rule 1 is on when the dialog opens, the opposite of the variable charts.
| Control | Default | Threshold box |
|---|---|---|
| Rule 1 - Beyond K sigma from center | on | none |
| Rule 2 - N consecutive points on same side of center | off | N = 9 |
| Rule 3 - N consecutive increasing or decreasing | off | N = 6 |
| Rule 4 - N consecutive alternating up/down | off | N = 14 |
Rule 1 here means "outside the drawn limits", and because the lower limit sits a hair above zero while intervals are almost always larger than that, in practice it fires only upward: a wait longer than the model expects.
There is no Rule 9 on this dialog. The zero-count Benneyan test belongs to the G Chart, where a count of zero is a meaningful value. Rules 5 through 8 are not offered either and are disabled in the engine for this chart.
Every threshold box accepts a whole number of 2 or more. A smaller value is refused, and unlike the G chart the T chart shows a message when it refuses.
The chart does not tell you which rule fired
Every rule you tick is tested, and the results are combined into a single verdict per point: this point broke at least one enabled rule, or it did not. Points that broke one are drawn in a different color from the rest, and that is the whole of the report. There is no per-rule label, no per-rule marker, no per-rule column, and no list of violations anywhere in the chart or the worksheet. A point that breaks four rules at once looks exactly like a point that breaks one, and the output never names either.
So when you turn several rules on, read a marked point as an instruction to go and look at the chart, not as a diagnosis. If you need to know which pattern caused a particular signal, the practical method is to switch the other rules off and rebuild the chart.
Display Options¶
| Control | Type | Default | What it does |
|---|---|---|---|
| Show 1-sigma / 2-sigma zone lines | check box | off | Draws the four intermediate quantile lines |
| Reference Lines: | grid with Value, Label, Color, Thickness, plus Add Reference Line and Apply | empty | Draws a horizontal line at a value you choose |
The zone lines are off by default. When you switch them on, read them for what they are on this chart: quantiles, not thirds of the band. The upper 1-sigma line sits much further from the center line than the lower one, because the fitted distribution is skewed. The formulas are on the Math Details page.
A reference line is drawn in the chart's own units, which are the units of your intervals, or the Time Unit you selected for a DateTime column.
Task Pane¶
The Control Chart pane that opens beside a finished chart has its own page: Task Pane.