Standard error and sample size

Work out how precise a sample mean is, and how big a sample a given precision needs.

What to build

A program that prints a standard error and a required sample size.

Requirements

  • Print standard error S (sigma ÷ √n, 3 decimal places) and sample size N, the smallest n whose standard error is at most the target.

Input

One line: sigma, n and the target error.

Output

Two lines.

Constraints

sigma and the target are more than 0; 1 ≤ n ≤ 10⁶.

Example

Input:

15 25 1

Output:

standard error 3.000
sample size 225

Why it works

The standard error shrinks with the square root of the sample size, so the size needed for a target grows with the square of sigma ÷ target.

Examples

Example 1 · Input
15 25 1
Expected output
standard error 3.000
sample size 225
Example 2 · Input
10 100 0.5
Expected output
standard error 1.000
sample size 400

Prompts inside input() are ignored when checking, so input("n: ") is fine.

Hints

Hint 1

The standard error is sigma ÷ √n.

Hint 2

The sample size for a target error is the smallest whole number at least (sigma ÷ target)²; subtract 0.000000001 before rounding up to absorb rounding errors.

Solution

Show one correct solution (try it yourself first)
from math import ceil

sigma, n, target = map(float, input().split())
print(f"standard error {sigma / n ** 0.5:.3f}")
print("sample size", ceil((sigma / target) ** 2 - 1e-9))

Lesson: Sampling and the central limit theorem →