Description
Apply fixed-count-per-tail winsorization retaining original observation order with a nonempty retained center; exact equal values use stable numeric order.
A descriptive statistic depends on what is counted, how observations are weighted, and which reference or denominator is used. Removing observations differs from replacing them at cutoffs, and a ratio of totals differs from a mean of individual ratios. Keep those conventions and measurement units explicit when comparing results.
Apply fixed-count-per-tail winsorization retaining original observation order with a nonempty retained center; exact equal values use stable numeric order.
When to use this tool
Use this tool to apply fixed-count-per-tail winsorization retaining original observation order with a nonempty retained center; exact equal values use stable numeric order.
How to use the tool and how it works
- Enter values, count using the field definitions below.
- Run the operation. Correct input errors before interpreting an output.
- Read result.values. Use the ordering, boundary rules and absence conditions described below.
Sort a copy, choose cutoffs at count and length − count − 1, and clamp each original value to those cutoffs without changing order or length.1
| Input field | Meaning and units | Accepted input |
|---|---|---|
| values | Finite observations in the measurement units used by reference or cutoffs. | Dense list of at most 128 finite numbers; integer/shape constraints above also apply. |
| count | Integer number of observations at each tail, not a percentage. | A finite safe integer within the operation’s stated bounds. |
| Output field | Interpretation |
|---|---|
| result | Object containing the named fields below. |
| result.values | Transformed or retained observations in the ordering described above. |
{
"input": {
"values": [
9,
1,
5,
2,
6
],
"count": 1
},
"output": {
"result": {
"values": [
6,
2,
5,
2,
6
]
}
}
}Limitations and assumptions
- Count must leave at least one central observation; winsorization replaces extreme values rather than deleting them.
- Use consistent units within each measured list and its reference/cutoffs. Ratios carry numerator units divided by denominator units unless those units cancel. Ordinary floating-point arithmetic is used; input validation does not establish statistical suitability.
- Every declared top-level input field is required and unexpected fields are rejected. Ordinary strings are limited to 8,192 UTF-16 code units and lists to 128 entries, with tighter operation-specific bounds where stated. JSON text has bounded nesting and node counts.
- Numeric outputs must be finite and no larger in absolute magnitude than the safe-integer limit; operations returning exact large integers use decimal text. Positions are zero-based and spans end exclusively unless the operation explicitly states a different convention.
Alternative or Complementary methods
Check the descriptive convention and denominator before comparing results. Use a separate inferential model when you need uncertainty, significance, causal interpretation or population estimates.
References
Similar or alternative tools
- Trim Fixed Count per Tail
Apply fixed-count-per-tail sorted trimming with a nonempty retained center; exact equal values use stable numeric order.
- Success Count Complement Table
Compute bounded observed success-count complements and rates, with null for zero-trial evidence; no inference of failure mechanisms or confidence intervals.