Home / Statistical Tools / Analysis Tools / Cpk/Histogram
Cpk/Histogram¶
Assess process capability by visualizing your data distribution against specification limits.
What is a Cpk/Histogram?¶
A Cpk/Histogram combines a histogram with process capability statistics. It shows your data distribution as bars, overlays normal curves (using within and overall standard deviation), and displays specification limit lines (USL, LSL, and optionally a Target value). Quantum XL calculates capability indices including Cp, Cpk, Pp, and Ppk, along with PPM (parts per million) defect rates.
The tool also supports distribution fitting for non-normal data (14 distributions compared, 11 selectable as the fitted distribution), data transformations (Box-Cox, Johnson), and confidence intervals for capability indices.
One tool, two ribbon buttons
The Capability Analysis and Histogram ribbon buttons open this same tool. Capability Analysis starts with the capability view (statistics and PDF curve, no bars), Histogram starts with the histogram view (bars, no curve), and either view can be changed afterward. The statistics and options are identical.
When to Use a Cpk/Histogram¶
- When you need to determine if a process meets its specification limits
- When you want to quantify process capability with Cpk and Ppk indices
- When you need to visualize where your data falls relative to USL and LSL
- When you want to assess the normality of your process data
- When you need to fit non-normal distributions and calculate capability from them
Example Output¶
A typical Cpk/Histogram shows bars representing the data distribution, red vertical lines at USL and LSL, overlaid normal curves, and a statistics table alongside with capability indices (Cpk, Ppk), PPM defect rates, and confidence intervals.
What the sheet records about your settings¶
The User input box in the top left of the worksheet does not only echo the columns you picked. Under a Settings heading it also states the options that produced the chart below it, so a sheet opened a year later says what it was built from rather than leaving you to guess which boxes were ticked.
This matters most where the choice leaves no other trace. Two Pareto charts drawn from the same data, one grouping the small categories into an "Others" bar and one dropping them, look like two different data sets unless the sheet says which was done.
Two rules decide what appears:
- Every selector and checkbox is printed on every run, whether or not you changed it from the default. A reader who sees nothing cannot tell a default from a setting the product does not report, so the box states all of them.
- A value you had to type is printed only when you typed it, and so is a setting that only means something once another box is ticked. Its absence is the answer.
The worksheet is protected¶
Under the last block in column A, in small italic text, the sheet carries one note:
This worksheet is protected to prevent accidental changes. To unprotect it, select the Review tab and then Unprotect Sheet.
Every analysis worksheet Quantum XL writes is protected, and every one carries this note. The wording is the same on all of them. Protection is there so that a click or a stray keystroke cannot overwrite a number the chart is drawn from, or move a chart off its anchor.
Three things are still allowed while the sheet is protected: formatting cells, formatting columns, and formatting rows. So you can widen a column, change a number format, or restyle a heading without unprotecting anything. Editing a value, editing a chart, and everything else are blocked.
For those, unprotect the sheet from Excel's Review tab, exactly as the note says. The protection carries no password, so Unprotect Sheet takes it straight off without asking for one.
This tool writes its rows straight into the User input box rather than under a Settings heading, so its labels end in a colon where every other tool's do not. The values shown below are the shapes each row takes, with your own numbers and column names in them.
| Row | What it reads |
|---|---|
| Subgrouping: | Subgroup Size = and the size, ID Column: and the column's name, or Subgroups Across Column Rows |
| Moving Range: | Average, Run Length = or Median, Run Length = and the length, or Mean Squared Successive Differences. Individuals data only |
| StDev Within Method: | R-bar, S-bar or Pooled StDev. Subgrouped data only, and not under the Boeing standard |
| Nelson Correction: | On or Off. Individuals data only, and not on the mean squared successive difference method |
| Unbiased Within:, Unbiased Overall: | each On or Off |
| Process Capability: | AIAG, Boeing D1-9000 or Custom |
| Sigma Tolerance: | the multiplier, printed even under a standard that fixes it |
| Cpk: | Sigma Within, Sigma Overall, Sigma Within & Between or Sigma Between |
| Ppk: | the same four. Not printed under Boeing D1-9000 |
| Confidence Level: | the level and its side, reading like 0.95 (Two-Sided), (Lower Bound) or (Upper Bound). Only when confidence intervals are on |
| Transformation: | None, Johnson, Box-Cox (optimal λ) when the fit chose the lambda, or Box-Cox (λ = ...) with the value when you typed one |
| Histogram Bins ( column name ): | the left point, bin count and bin width for that column, each a number, Auto, or a bare Manual for a box switched to manual and left empty. One row per measurement column |
| Distribution Fit ( column name ): | the fitted distribution's name. One row per column, only when the curve is drawn |
Learn More¶
- How-To: Step-by-step tutorials for different scenarios
- Preparing Your Data: Data types, specification limits, and subgroup options
- Options: Capability settings, chart options, distribution fitting, and more
- Math Details: The exact formulas behind every statistic