Home / DOE / Charts / Chart Source
Chart Source¶
Every DOE chart is built from one worksheet, and which kind of worksheet it can use decides what the chart is able to say. This page explains the two sources, which charts offer which, and the rules that follow from the choice.
The two sources¶
Some charts can be built from either kind of sheet, and the two answer different questions.
A regression sheet gives the fitted model's prediction: a smooth value at any combination of settings, including combinations no run used. A design sheet gives the raw data: the average of the responses actually measured at each setting, and where a setting has replicates, their standard deviation. The design sheet can only speak about settings that were run.
Where both are offered, the dialog lists every eligible worksheet in the workbook and you pick one.
Which charts offer which¶
| Chart | Regression sheet | Design sheet |
|---|---|---|
| Pareto of Regression Coefficients | yes | no |
| Interaction Plots | yes | yes, quantitative outputs only |
| Main Effects Plots | yes | yes, quantitative outputs only |
| Thumbnail Plot | yes | yes, quantitative outputs only |
| Cube Plot | yes | yes, quantitative outputs only |
| Surface Plot | yes | no |
| Contour Plot | yes | no |
| Trellis | yes | no |
| Observed vs Predicted | yes | no |
| Residual Plots | yes | no |
The quantitative-only rule is not a Cube Plot special case. Whenever the source is a design sheet, only a quantitative output is offered, for every one of the four charts that accept one. A design sheet whose outputs are all binary or nominal does not appear in the Outputs tree at all.
The pattern behind the table is simple. A chart that needs a value at a setting nobody ran can only come from a model, so it is regression only. A chart that reports what was measured can come from either.
Surface, Contour and Trellis draw a value at every point of a continuous region, most of which was never run. Observed vs Predicted and Residual Plots need a fitted value for each run. A Pareto is a chart of coefficients, and only a regression produces coefficients.
Reading the numbers a chart shows¶
A chart built from a regression sheet does not re-fit the model. It reads the coefficients already on that sheet and evaluates the prediction equation with them.
Two things follow. The chart always agrees with the sheet it came from, so if you edit a coefficient the next chart changes with it. And a chart is only as current as its regression sheet: change the design and the old charts do not update, because nothing re-runs the regression for them.
A term that is not in the regression table, or that is switched off for the output level being charted, contributes exactly zero to the prediction. It is not an error and there is no warning: the chart is simply drawn from the reduced model. So a chart reflects which terms are active at the moment it is created, and two charts of the same output can differ because the active terms changed between them.
From raw run data¶
From a design sheet the two plotted values come straight from the run data. The response value is the average of the responses at each distinct setting of the charted factors. The variation value is the average, over that same setting, of the standard deviations within each replicate group, using the sample standard deviation.
A setting with no replicate group of two or more runs has no variation to report and is left blank rather than shown as zero.
What the sheet records about its source¶
Each chart sheet names the sheet it came from in its heading rows, so a chart you find later can be traced back to the model or the data behind it. A Pareto names both the design sheet and the regression sheet; the others name the one source they used.