pystatpower.proportion.single.noninferiority ¶

FUNCTION	DESCRIPTION
`solve_power`	Calculate the statistical power.
`solve_size`	Estimate the required sample size.
`solve_proportion`	Estimate the required proportion under the alternative hypothesis.
`solve_null_proportion`	Estimate the required proportion under the null hypothesis.
`solve_noninferiority_proportion`	Estimate the required non-inferiority proportion.
`solve_margin`	Estimate the required non-inferiority margin.

solve_power ¶

solve_power(
    *,
    proportion: float,
    null_proportion: float | None = None,
    margin: float | None = None,
    noninferiority_proportion: float | None = None,
    size: int,
    alternative: Literal["greater", "less"],
    alpha: float = 0.025,
    method: Literal["z-p0", "z-phat"] = "z-phat",
    continuity_correction: bool,
) -> float

Calculate the statistical power.

PARAMETER	DESCRIPTION
`proportion`	Proportion under the alternative hypothesis. TYPE: `float`
`null_proportion`	Proportion under the null hypothesis. Ignored if `noninferiority_proportion` is specified; otherwise, required along with `margin`. TYPE: `float \| None` DEFAULT: `None`
`margin`	The non-inferiority margin. Ignored if `noninferiority_proportion` is specified; otherwise, required along with `null_proportion`. Tip Regardless of whether `alternative` is specified as `'greater'` or `'less'`, you can always specify this parameter as either positive or negative, as you prefer. Internally, the value of `margin` is converted before actual calculation takes place. If `alternative` is `'greater'`, the internally used margin is `-abs(margin)`. If `alternative` is `'less'`, the internally used margin is `abs(margin)`. TYPE: `float \| None` DEFAULT: `None`
`noninferiority_proportion`	The non-inferiority proportion. Required if either `null_proportion` or `margin` is omitted. TYPE: `float \| None` DEFAULT: `None`
`size`	Sample size. TYPE: `int`
`alternative`	Type of the alternative hypothesis: If `alternative` is `'greater'`, the alternative hypothesis is \(p - p_0 > \delta \ (\delta < 0)\) If `alternative` is `'less'`, the alternative hypothesis is \(p - p_0 < \delta \ (\delta > 0)\) TYPE: `Literal['greater', 'less']`
`alpha`	Significance level. The non-inferiority test is a one-sided test, with a significance level of 0.025 being commonly used. TYPE: `float` DEFAULT: `0.025`
`method`	The method used to construct the test statistic. `'z-p0'`: Standard normal distribution (large sample approximation), using p0 to calculate the variance. `'z-phat'`: Standard normal distribution (large sample approximation), using phat to calculate the variance. TYPE: `Literal['z-p0', 'z-phat']` DEFAULT: `'z-phat'`
`continuity_correction`	Whether to apply the continuity correction. TYPE: `bool`

RETURNS	DESCRIPTION
`float`	The statistical power of the test.

RAISES	DESCRIPTION
`ValueError`	If `noninferiority_proportion` is omitted, and either `null_proportion` or `margin` is missing.

solve_size ¶

solve_size(
    *,
    proportion: float,
    null_proportion: float | None = None,
    margin: float | None = None,
    noninferiority_proportion: float | None = None,
    alternative: Literal["greater", "less"],
    alpha: float = 0.025,
    power: float = 0.8,
    method: Literal["z-p0", "z-phat"] = "z-phat",
    continuity_correction: bool = False,
) -> int

Estimate the required sample size.

PARAMETER	DESCRIPTION
`proportion`	Proportion under the alternative hypothesis. TYPE: `float`
`null_proportion`	Proportion under the null hypothesis. Ignored if `noninferiority_proportion` is specified; otherwise, required along with `margin`. TYPE: `float \| None` DEFAULT: `None`
`margin`	The non-inferiority margin. Ignored if `noninferiority_proportion` is specified; otherwise, required along with `null_proportion`. Tip Regardless of whether `alternative` is specified as `'greater'` or `'less'`, you can always specify this parameter as either positive or negative, as you prefer. Internally, the value of `margin` is converted before actual calculation takes place. If `alternative` is `'greater'`, the internally used margin is `-abs(margin)`. If `alternative` is `'less'`, the internally used margin is `abs(margin)`. TYPE: `float \| None` DEFAULT: `None`
`noninferiority_proportion`	The non-inferiority proportion. Required if either `null_proportion` or `margin` is omitted. TYPE: `float \| None` DEFAULT: `None`
`alternative`	Type of the alternative hypothesis: If `alternative` is `'greater'`, the alternative hypothesis is \(p - p_0 > \delta \ (\delta < 0)\) If `alternative` is `'less'`, the alternative hypothesis is \(p - p_0 < \delta \ (\delta > 0)\) TYPE: `Literal['greater', 'less']`
`alpha`	Significance level. The non-inferiority test is a one-sided test, with a significance level of 0.025 being commonly used. TYPE: `float` DEFAULT: `0.025`
`power`	Expected statistical power. 0.8 is a commonly used value for statistical power. TYPE: `float` DEFAULT: `0.8`
`method`	The method used to construct the test statistic. `'z-p0'`: Standard normal distribution (large sample approximation), using p0 to calculate the variance. `'z-phat'`: Standard normal distribution (large sample approximation), using phat to calculate the variance. TYPE: `Literal['z-p0', 'z-phat']` DEFAULT: `'z-phat'`
`continuity_correction`	Whether to apply the continuity correction. TYPE: `bool` DEFAULT: `False`

RETURNS	DESCRIPTION
`int`	The required sample size.

RAISES	DESCRIPTION
`ValueError`	If `noninferiority_proportion` is omitted, and either `null_proportion` or `margin` is missing.

solve_proportion ¶

solve_proportion(
    *,
    null_proportion: float | None = None,
    margin: float | None = None,
    noninferiority_proportion: float | None = None,
    size: int,
    alternative: Literal["greater", "less"],
    alpha: float = 0.025,
    power: float = 0.8,
    method: Literal["z-p0", "z-phat"] = "z-phat",
    continuity_correction: bool = False,
) -> float

Estimate the required proportion under the alternative hypothesis.

PARAMETER	DESCRIPTION
`null_proportion`	Proportion under the null hypothesis. Ignored if `noninferiority_proportion` is specified; otherwise, required along with `margin`. TYPE: `float \| None` DEFAULT: `None`
`margin`	The non-inferiority margin. Ignored if `noninferiority_proportion` is specified; otherwise, required along with `margin`. Tip Regardless of whether `alternative` is specified as `'greater'` or `'less'`, you can always specify this parameter as either positive or negative, as you prefer. Internally, the value of `margin` is converted before actual calculation takes place. If `alternative` is `'greater'`, the internally used margin is `-abs(margin)`. If `alternative` is `'less'`, the internally used margin is `abs(margin)`. TYPE: `float \| None` DEFAULT: `None`
`noninferiority_proportion`	The non-inferiority proportion. Required if either `null_proportion` or `margin` is omitted. TYPE: `float \| None` DEFAULT: `None`
`size`	Sample size. TYPE: `int`
`alternative`	Type of the alternative hypothesis: If `alternative` is `'greater'`, the alternative hypothesis is \(p - p_0 > \delta \ (\delta < 0)\) If `alternative` is `'less'`, the alternative hypothesis is \(p - p_0 < \delta \ (\delta > 0)\) TYPE: `Literal['greater', 'less']`
`alpha`	Significance level. The non-inferiority test is a one-sided test, with a significance level of 0.025 being commonly used. TYPE: `float` DEFAULT: `0.025`
`power`	Expected statistical power. 0.8 is a commonly used value for statistical power. TYPE: `float` DEFAULT: `0.8`
`method`	The method used to construct the test statistic. `'z-p0'`: Standard normal distribution (large sample approximation), using p0 to calculate the variance. `'z-phat'`: Standard normal distribution (large sample approximation), using phat to calculate the variance. TYPE: `Literal['z-p0', 'z-phat']` DEFAULT: `'z-phat'`
`continuity_correction`	Whether to apply the continuity correction. TYPE: `bool` DEFAULT: `False`

RETURNS	DESCRIPTION
`float`	The required proportion under the alternative hypothesis.

RAISES	DESCRIPTION
`ValueError`	If `noninferiority_proportion` is omitted, and either `null_proportion` or `margin` is missing.

Notes

The value range of the alternative hypothesis proportion \(p\) is determined by the null hypothesis proportion \(p_0\) and the non-inferiority margin \(\delta\).

If \(\delta < 0\), we have:

\[ \begin{cases} p > p_0 + \delta \\ 0 < p < 1 \end{cases} \Rightarrow \operatorname{max}(p_0 + \delta, 0) < p < 1 \]

If \(\delta > 0\), we have:

\[ \begin{cases} p < p_0 + \delta \\ 0 < p < 1 \end{cases} \Rightarrow 0 < p_0 < \operatorname{min}(p_0 + \delta, 1) \]

solve_null_proportion ¶

solve_null_proportion(
    *,
    proportion: float,
    margin: float,
    size: int,
    alternative: Literal["greater", "less"],
    alpha: float = 0.025,
    power: float = 0.8,
    method: Literal["z-p0", "z-phat"] = "z-phat",
    continuity_correction: bool = False,
) -> float

Estimate the required proportion under the null hypothesis.

PARAMETER	DESCRIPTION
`proportion`	Proportion under the alternative hypothesis. TYPE: `float`
`margin`	The non-inferiority margin. Tip Regardless of whether `alternative` is specified as `'greater'` or `'less'`, you can always specify this parameter as either positive or negative, as you prefer. Internally, the value of `margin` is converted before actual calculation takes place. If `alternative` is `'greater'`, the internally used margin is `-abs(margin)`. If `alternative` is `'less'`, the internally used margin is `abs(margin)`. TYPE: `float`
`size`	Sample size. TYPE: `int`
`alternative`	Type of the alternative hypothesis: If `alternative` is `'greater'`, the alternative hypothesis is \(p - p_0 > \delta \ (\delta < 0)\) If `alternative` is `'less'`, the alternative hypothesis is \(p - p_0 < \delta \ (\delta > 0)\) TYPE: `Literal['greater', 'less']`
`alpha`	Significance level. The non-inferiority test is a one-sided test, with a significance level of 0.025 being commonly used. TYPE: `float` DEFAULT: `0.025`
`power`	Expected statistical power. 0.8 is a commonly used value for statistical power. TYPE: `float` DEFAULT: `0.8`
`method`	The method used to construct the test statistic. `'z-p0'`: Standard normal distribution (large sample approximation), using p0 to calculate the variance. `'z-phat'`: Standard normal distribution (large sample approximation), using phat to calculate the variance. TYPE: `Literal['z-p0', 'z-phat']` DEFAULT: `'z-phat'`
`continuity_correction`	Whether to apply the continuity correction. TYPE: `bool` DEFAULT: `False`

RETURNS	DESCRIPTION
`float`	The required proportion under the null hypothesis.

Notes

The value range of the null hypothesis proportion \(p_0\) is determined by the alternative hypothesis proportion \(p\) and the non-inferiority margin \(\delta\).

If \(\delta < 0\), we have:

\[ \begin{cases} p_0 < p - \delta \\ 0 < p_0 < 1 \\ 0 < p_0 + \delta < 1 \end{cases} \Rightarrow -\delta < p_0 < \operatorname{min}(p - \delta, 1) \]

If \(\delta > 0\), we have:

\[ \begin{cases} p_0 > p - \delta \\ 0 < p_0 < 1 \\ 0 < p_0 + \delta < 1 \end{cases} \Rightarrow \operatorname{max}(p - \delta, 0) < p_0 < 1 - \delta \]

solve_noninferiority_proportion ¶

solve_noninferiority_proportion(
    *,
    proportion: float,
    size: int,
    alternative: Literal["greater", "less"],
    alpha: float = 0.025,
    power: float = 0.8,
    method: Literal["z-p0", "z-phat"] = "z-phat",
    continuity_correction: bool = False,
) -> float

Estimate the required non-inferiority proportion.

PARAMETER	DESCRIPTION
`proportion`	Proportion under the alternative hypothesis. TYPE: `float`
`size`	Sample size. TYPE: `int`
`alternative`	Type of the alternative hypothesis: If `alternative` is `'greater'`, the alternative hypothesis is \(p - p_0 > \delta \ (\delta < 0)\) If `alternative` is `'less'`, the alternative hypothesis is \(p - p_0 < \delta \ (\delta > 0)\) TYPE: `Literal['greater', 'less']`
`alpha`	Significance level. The non-inferiority test is a one-sided test, with a significance level of 0.025 being commonly used. TYPE: `float` DEFAULT: `0.025`
`power`	Expected statistical power. 0.8 is a commonly used value for statistical power. TYPE: `float` DEFAULT: `0.8`
`method`	The method used to construct the test statistic. `'z-p0'`: Standard normal distribution (large sample approximation), using p0 to calculate the variance. `'z-phat'`: Standard normal distribution (large sample approximation), using phat to calculate the variance. TYPE: `Literal['z-p0', 'z-phat']` DEFAULT: `'z-phat'`
`continuity_correction`	Whether to apply the continuity correction. TYPE: `bool` DEFAULT: `False`

RETURNS	DESCRIPTION
`float`	The required non-inferiority proportion.

Notes

The value range of the non-inferiority hypothesis proportion \(p_{\text{noninf}}\) is determined by the alternative hypothesis proportion \(p\).

If alternative is 'greater', that is, higher proportions are better, we have:

\[ \begin{cases} p_{\text{noninf}} < p \\ 0 < p_{\text{noninf}} < 1 \\ \end{cases} \ \Rightarrow 0 < p_{\text{noninf}} < p \]

If alternative is 'less', that is, higher proportions are worse, we have:

\[ \begin{cases} p_{\text{noninf}} > p \\ 0 < p_{\text{noninf}} < 1 \\ \end{cases} \ \Rightarrow p < p_{\text{noninf}} < 1 \]

solve_margin ¶

solve_margin(
    *,
    proportion: float,
    null_proportion: float,
    size: int,
    alternative: Literal["greater", "less"],
    alpha: float = 0.025,
    power: float = 0.8,
    method: Literal["z-p0", "z-phat"] = "z-phat",
    continuity_correction: bool = False,
) -> float

Estimate the required non-inferiority margin.

PARAMETER	DESCRIPTION
`proportion`	Proportion under the alternative hypothesis. TYPE: `float`
`null_proportion`	Proportion under the null hypothesis. TYPE: `float`
`size`	Sample size. TYPE: `int`
`alternative`	Type of the alternative hypothesis: If `alternative` is `'greater'`, the alternative hypothesis is \(p - p_0 > \delta \ (\delta < 0)\) If `alternative` is `'less'`, the alternative hypothesis is \(p - p_0 < \delta \ (\delta > 0)\) TYPE: `Literal['greater', 'less']`
`alpha`	Significance level. The non-inferiority test is a one-sided test, with a significance level of 0.025 being commonly used. TYPE: `float` DEFAULT: `0.025`
`power`	Expected statistical power. 0.8 is a commonly used value for statistical power. TYPE: `float` DEFAULT: `0.8`
`method`	The method used to construct the test statistic. `'z-p0'`: Standard normal distribution (large sample approximation), using p0 to calculate the variance. `'z-phat'`: Standard normal distribution (large sample approximation), using phat to calculate the variance. TYPE: `Literal['z-p0', 'z-phat']` DEFAULT: `'z-phat'`
`continuity_correction`	Whether to apply the continuity correction. TYPE: `bool` DEFAULT: `False`

RETURNS	DESCRIPTION
`float`	The required non-inferiority margin.

Notes

The value range of the non-inferiority margin \(\delta\) is determined by the alternative hypothesis proportion \(p\) and the null hypothesis proportion \(p_0\).

If \(\delta < 0\), we have:

\[ \begin{cases} \delta < p - p_0 \\ 0 < p_0 + \delta < 1 \\ \delta < 0 \end{cases} \Rightarrow -p_0 < \delta < \operatorname{min}(p - p_0, 0) \]

If \(\delta > 0\), we have:

\[ \begin{cases} \delta > p - p_0 \\ 0 < p_0 + \delta < 1 \\ \delta > 0 \end{cases} \Rightarrow \operatorname{max}(p - p_0, 0) < \delta < 1 - p_0 \]