pystatpower.correlation.inequality ¶

Functions:

Name	Description
`solve_power`	Calculate the statistical power of the difference test between two correlation coefficients.
`solve_size`	Estimate the required sample size for the difference test between two correlation coefficients.
`solve_correlation`	Estimate the required correlation coefficient under the alternative hypothesis for the difference test between two correlation coefficients.
`solve_null_correlation`	Estimete the required correlation coefficient under the null hypothesis for the difference test between two correlation coefficients.

solve_power ¶

solve_power(
    *,
    null_correlation: float,
    correlation: float,
    size: int,
    alternative: Literal[
        "two-sided", "greater", "less"
    ] = "two-sided",
    alpha: float = 0.05,
    bias_adj: bool = False,
) -> float

Calculate the statistical power of the difference test between two correlation coefficients.

Parameters:

Name	Type	Description	Default
`null_correlation`	`float`	Correlation coefficient under the null hypothesis.	required
`correlation`	`float`	Correlation coefficient under the alternative hypothesis.	required
`size`	`int`	Sample size.	required
`alternative`	`Literal['two-sided', 'greater', 'less']`	Type of alternative hypothesis. `'two-sided'`: Two-sided alternative hypothesis: \(\rho_1 \neq \rho_0\) `'greater'`: Upper one-sided alternative hypothesis: \(\rho_1 > \rho_0\) `'less'`: Lower one-sided alternative hypothesis: \(\rho_1 < \rho_0\)	`'two-sided'`
`alpha`	`float`	Significance level. If `alternative` is `'two-sided'`, provide the two-sided significance level. If `alternative` is `'greater'` or `'less'`, provide the one-sided significance level.	`0.05`
`bias_adj`	`bool`	Specify whether the bias adjustment is used or not.	`False`

Returns:

Type	Description
`float`	Power of the test.

solve_size ¶

solve_size(
    *,
    null_correlation: float,
    correlation: float,
    alternative: Literal[
        "two-sided", "greater", "less"
    ] = "two-sided",
    alpha: float = 0.05,
    power: float = 0.8,
    bias_adj: bool = False,
) -> int

Estimate the required sample size for the difference test between two correlation coefficients.

Parameters:

Name	Type	Description	Default
`null_correlation`	`float`	Correlation coefficient under the null hypothesis.	required
`correlation`	`float`	Correlation coefficient under the alternative hypothesis.	required
`alternative`	`Literal['two-sided', 'greater', 'less']`	Type of alternative hypothesis. `'two-sided'`: Two-sided alternative hypothesis: \(\rho_1 \neq \rho_0\) `'greater'`: Upper one-sided alternative hypothesis: \(\rho_1 > \rho_0\) `'less'`: Lower one-sided alternative hypothesis: \(\rho_1 < \rho_0\)	`'two-sided'`
`alpha`	`float`	Significance level. If `alternative` is `'two-sided'`, provide the two-sided significance level. If `alternative` is `'greater'` or `'less'`, provide the one-sided significance level.	`0.05`
`power`	`float`	Power of the test.	`0.8`
`bias_adj`	`bool`	Specify whether the bias adjustment is used or not.	`False`

Returns:

Type	Description
`int`	The required sample size.

solve_correlation ¶

solve_correlation(
    *,
    null_correlation: float,
    size: int,
    alternative: Literal[
        "two-sided", "greater", "less"
    ] = "two-sided",
    alpha: float = 0.05,
    power: float = 0.8,
    bias_adj: bool = False,
    search_direction: Literal["above", "below"] = "above",
) -> float

Estimate the required correlation coefficient under the alternative hypothesis for the difference test between two correlation coefficients.

Parameters:

Name	Type	Description	Default
`null_correlation`	`float`	Correlation coefficient under the null hypothesis.	required
`size`	`int`	Sample size.	required
`alternative`	`Literal['two-sided', 'greater', 'less']`	Type of alternative hypothesis. `'two-sided'`: Two-sided alternative hypothesis: \(\rho_1 \neq \rho_0\) `'greater'`: Upper one-sided alternative hypothesis: \(\rho_1 > \rho_0\) `'less'`: Lower one-sided alternative hypothesis: \(\rho_1 < \rho_0\)	`'two-sided'`
`alpha`	`float`	Significance level. If `alternative` is `'two-sided'`, provide the two-sided significance level. If `alternative` is `'greater'` or `'less'`, provide the one-sided significance level.	`0.05`
`power`	`float`	Power of the test.	`0.8`
`bias_adj`	`bool`	Specify whether the bias adjustment is used or not.	`False`
`search_direction`	`Literal['above', 'below']`	Specify whether to search for the alternative correlation below or above the null correlation.	`'above'`

Returns:

Type	Description
`float`	required correlation coefficient under the alternative hypothesis.

solve_null_correlation ¶

solve_null_correlation(
    *,
    correlation: float,
    size: int,
    alternative: Literal[
        "two-sided", "greater", "less"
    ] = "two-sided",
    alpha: float = 0.05,
    power: float = 0.8,
    bias_adj: bool = False,
    search_direction: Literal["above", "below"] = "below",
) -> float

Estimete the required correlation coefficient under the null hypothesis for the difference test between two correlation coefficients.

Parameters:

Name	Type	Description	Default
`correlation`	`float`	Correlation coefficient under the alternative hypothesis.	required
`size`	`int`	Sample size.	required
`alternative`	`Literal['two-sided', 'greater', 'less']`	Type of alternative hypothesis. `'two-sided'`: Two-sided alternative hypothesis: \(\rho_1 \neq \rho_0\) `'greater'`: Upper one-sided alternative hypothesis: \(\rho_1 > \rho_0\) `'less'`: Lower one-sided alternative hypothesis: \(\rho_1 < \rho_0\)	`'two-sided'`
`alpha`	`float`	Two-sided significance level. If `alternative` is `'two-sided'`, provide the two-sided significance level. If `alternative` is `'less'` or `'greater'`, provide the one-sided significance level.	`0.05`
`power`	`float`	Power of the test.	`0.8`
`bias_adj`	`bool`	Specify whether the bias adjustment is used or not.	`False`
`search_direction`	`Literal['above', 'below']`	Specify whether to search for the null correlation below or above the alternative correlation.	`'below'`

Returns:

Type	Description
`float`	The required correlation coefficient under the null hypothesis.