Home / Statistical Tools / Analysis Tools / Cpkhistogram / Preparing Your Data
Preparing Your Data¶
Quantum XL is flexible about data layout. This page explains how different data types behave in Cpk histograms and helps you prepare your data correctly.
Subgroup Options
By default, Quantum XL treats each row as an individual observation (subgroup size = 1). You can also define subgroups by specifying a subgroup size (consecutive rows grouped together), an ID column (rows with the same ID form a subgroup), or by selecting "Subgroups Across Column Rows" where each row across all selected columns forms one subgroup — the subgroup size equals the number of columns.
How Data Types Work¶
The table below shows what happens when you assign each column type in the Cpk/Histogram dialog. Click a numbered badge to see an example using that configuration.
| Data Column | ||||||
|---|---|---|---|---|---|---|
| Nominal | Continuous | Integer | Count | DateTime | ||
| Frequency Column | None | Not Allowed | Each value is one observation. Histogram bins computed automatically. Spec limits (USL/LSL) entered per column. 12 | Each value is one observation. Histogram bins computed automatically. Spec limits (USL/LSL) entered per column. | Each value is one observation. Histogram bins computed automatically. Spec limits (USL/LSL) entered per column. | Not Allowed |
| Nominal | — | |||||
| Continuous | — | |||||
| Integer | — | |||||
| Count | — | |||||
| DateTime | — | |||||
— indicates this frequency type is not available for selection. Cpk/Histogram does not support frequency columns.
Examples¶
1 Quick Start — Single column with specification limits
2 Multiple Data Columns — Multiple columns with different spec limits
GroupBy — Compare capability across groups
Data Layout Options¶
Individual Measurements¶
The most common layout. Each row is one observation.
| Diameter |
|---|
| 9.82 |
| 10.05 |
| 9.97 |
| 10.12 |
Each row is one data point. Enter the specification limits (USL, LSL, and optionally Target) in the Data tab for each column.
Subgrouped Data¶
When your data is organized in subgroups (e.g., 5 consecutive measurements form one subgroup), set the subgroup size in the Data tab. This affects how the within-subgroup standard deviation is estimated, which in turn affects Cpk calculations.
| Measurement |
|---|
| 9.82 |
| 10.05 |
| 9.97 |
| 10.12 |
| 9.91 |
With a subgroup size of 5, these 5 rows form one subgroup. The within-subgroup standard deviation is estimated from variation within subgroups, while the overall standard deviation uses all data.
Common Mistakes¶
Avoid These Issues
- Forgetting to enter specification limits — Without USL and/or LSL, Quantum XL cannot calculate capability indices (Cp, Cpk, Pp, Ppk) or PPM defect rates. The histogram will still be created, but without capability statistics.
- Selecting Nominal or DateTime columns — Cpk/Histogram requires numeric data (Continuous, Integer, or Count). Nominal and DateTime columns cannot be used as data columns.
- Too few data points — Capability statistics require sufficient data to be meaningful. Very small samples produce unreliable Cpk/Ppk estimates.
- Missing data — Empty cells in data columns are excluded from the analysis.