mbapy.stats.test

This module provides statistical analysis functions for data analysis.

Functions

get_interval(mean = None, se = None, data:Optional[Union[np.ndarray, List[int], pd.Series]] = None, confidence:float = 0.95) -> Tuple[Tuple[float, float], Dict[str, Any]]

Calculate the confidence interval.

Parameters

mean (float, optional): The mean value. Defaults to None.
se (float, optional): The standard error. Defaults to None.
data (np.ndarray or List[int] or pd.Series, optional): The data. Defaults to None.
confidence (float, optional): The confidence level. Defaults to 0.95.

Returns

Tuple[Tuple[float, float], Dict[str, Any]]: The confidence interval and additional information.

Notes

The confidence interval is calculated as ± 1.96 * SE or other depends on the confidence level.

Example

get_interval(mean=10, se=2, confidence=0.95)

ttest_1samp(x1 = None, x2:float = None, factors:Dict[str, List[str]] = None, tag:str = None, df:pd.DataFrame = None, **kwargs) -> Tuple[float, float]

Perform a one-sample t-test.

Parameters

x1 (float or None, optional): The first sample. Defaults to None.
x2 (float, optional): The second sample. Defaults to None.
factors (Dict[str, List[str]], optional): The factors for data filtering. Defaults to None.
tag (str, optional): The column name of the data. Defaults to None.
df (pd.DataFrame, optional): The data frame. Defaults to None.
kwargs: Additional arguments for the t-test.

Returns

Tuple[float, float]: The p-value and the t-statistic.

Example

ttest_1samp(x1=10, tag='score', df=data)

ttest_ind(x1 = None, x2 = None, factors:Dict[str, List[str]] = None, tag:str = None, df:pd.DataFrame = None, **kwargs) -> Tuple[float, float]

Perform an independent samples t-test.

Parameters

x1 (float or None, optional): The first sample. Defaults to None.
x2 (float or None, optional): The second sample. Defaults to None.
factors (Dict[str, List[str]], optional): The factors for data filtering. Defaults to None.
tag (str, optional): The column name of the data. Defaults to None.
df (pd.DataFrame, optional): The data frame. Defaults to None.
kwargs: Additional arguments for the t-test.

Returns

Tuple[float, float]: The p-value from the Levene's test and the p-value from the t-test.

Example

ttest_ind(factors={'group': ['A', 'B']}, tag='score', df=data)

ttest_rel(x1 = None, x2 = None, factors:Dict[str, List[str]] = None, tag:str = None, df:pd.DataFrame = None, **kwargs) -> Tuple[float, float]

Perform a paired samples t-test.

Parameters

x1 (float or None, optional): The first sample. Defaults to None.
x2 (float or None, optional): The second sample. Defaults to None.
factors (Dict[str, List[str]], optional): The factors for data filtering. Defaults to None.
tag (str, optional): The column name of the data. Defaults to None.
df (pd.DataFrame, optional): The data frame. Defaults to None.
kwargs: Additional arguments for the t-test.

Returns

Tuple[float, float]: The p-value and the t-statistic.

Example

ttest_rel(factors={'group': ['A', 'B']}, tag='score', df=data)

mannwhitneyu(x1 = None, x2 = None, factors:Dict[str, List[str]] = None, tag:str = None, df:pd.DataFrame = None, **kwargs) -> Tuple[float, float]

Perform a Mann-Whitney U test.

Parameters

x1 (float or None, optional): The first sample. Defaults to None.
x2 (float or None, optional): The second sample. Defaults to None.
factors (Dict[str, List[str]], optional): The factors for data filtering. Defaults to None.
tag (str, optional): The column name of the data. Defaults to None.
df (pd.DataFrame, optional): The data frame. Defaults to None.
kwargs: Additional arguments for the Mann-Whitney U test.

Returns

Tuple[float, float]: The p-value and the U statistic.

Example

mannwhitneyu(factors={'group': ['A', 'B']}, tag='score', df=data)

shapiro(x1 = None, factors:Dict[str, List[str]] = None, tag:str = None, df:pd.DataFrame = None, **kwargs) -> Tuple[float, float]

Perform a Shapiro-Wilk test for normality.

Parameters

x1 (float or None, optional): The sample. Defaults to None.
factors (Dict[str, List[str]], optional): The factors for data filtering. Defaults to None.
tag (str, optional): The column name of the data. Defaults to None.
df (pd.DataFrame, optional): The data frame. Defaults to None.
kwargs: Additional arguments for the Shapiro-Wilk test.

Returns

Tuple[float, float]: The test statistic and the p-value.

Example

shapiro(factors={'group': ['A', 'B']}, tag='score', df=data)

pearsonr(x1 = None, x2 = None, factors:Dict[str, List[str]] = None, tags:List[str] = None, df:pd.DataFrame = None, **kwargs) -> Tuple[float, float]

Calculate the Pearson correlation coefficient.

Parameters

x1 (float or None, optional): The first sample. Defaults to None.
x2 (float or None, optional): The second sample. Defaults to None.
factors (Dict[str, List[str]], optional): The factors for data filtering. Defaults to None.
tags (List[str], optional): The column names of the data. Defaults to None.
df (pd.DataFrame, optional): The data frame. Defaults to None.
kwargs: Additional arguments for the Pearson correlation coefficient.

Returns

Tuple[float, float]: The correlation coefficient and the p-value.

Example

pearsonr(factors={'group': ['A', 'B']}, tags=['x', 'y'], df=data)

chi2_contingency(observed = None, factors:Dict[str, List[str]] = None, tag:str = None, df:pd.DataFrame = None, **kwargs) -> Tuple[Tuple[float, float, int, np.ndarray], pd.DataFrame]

Perform a chi-squared test for independence.

Parameters

observed (np.ndarray or None, optional): The observed frequencies. Defaults to None.
factors (Dict[str, List[str]], optional): The factors for data filtering. Defaults to None.
tag (str, optional): The column name of the data. Defaults to None.
df (pd.DataFrame, optional): The data frame. Defaults to None.
kwargs: Additional arguments for the chi-squared test.

Returns

Tuple[Tuple[float, float, int, np.ndarray], pd.DataFrame]: The test statistic, the p-value, the degrees of freedom, the expected frequencies, and the observed frequencies.

Example

chi2_contingency(factors={'group': ['A', 'B']}, tag='outcome', df=data)

fisher_exact(observed = None, factors:Dict[str, List[str]] = None, tag:str = None, df:pd.DataFrame = None, **kwargs) -> Tuple[Tuple[float, float], pd.DataFrame]

Perform a Fisher's exact test.

Parameters

observed (np.ndarray or None, optional): The observed frequencies. Defaults to None.
factors (Dict[str, List[str]], optional): The factors for data filtering. Defaults to None.
tag (str, optional): The column name of the data. Defaults to None.
df (pd.DataFrame, optional): The data frame. Defaults to None.
kwargs: Additional arguments for the Fisher's exact test.

Returns

Tuple[Tuple[float, float], pd.DataFrame]: The odds ratio, the p-value, the expected frequencies, and the observed frequencies.

Example

fisher_exact(factors={'group': ['A', 'B']}, tag='outcome', df=data)

f_oneway(Xs:list = None, factors:Dict[str, List[str]] = None, tag:str = None, df:pd.DataFrame = None) -> Tuple[float, float]

Perform a one-way ANOVA.

Parameters

Xs (list or None, optional): The samples. Defaults to None.
factors (Dict[str, List[str]], optional): The factors for data filtering. Defaults to None.
tag (str, optional): The column name of the data. Defaults to None.
df (pd.DataFrame, optional): The data frame. Defaults to None.

Returns

Tuple[float, float]: The F-statistic and the p-value.

Example

f_oneway(Xs=[x1, x2, x3], factors={'group': ['A', 'B', 'C']}, tag='score', df=data)

multicomp_turkeyHSD(factors:Dict[str, List[str]], tag:str, df:pd.DataFrame, alpha:float = 0.05) -> sm.stats.multicomp.pairwise_tukeyhsd

Perform a Tukey's HSD test for multiple comparisons.

Parameters

factors (Dict[str, List[str]]): The factors for data filtering.
tag (str): The column name of the data.
df (pd.DataFrame): The data frame.
alpha (float, optional): The significance level. Defaults to 0.05.

Returns

sm.stats.multicomp.pairwise_tukeyhsd: The result object from the Tukey's HSD test.

Example

multicomp_turkeyHSD(factors={'group': ['A', 'B', 'C']}, tag='score', df=data)

turkey_to_table(turkey_result:sm.stats.multicomp.pairwise_tukeyhsd) -> pd.DataFrame

Generate a table summarizing the results of the Tukey's HSD test.

Parameters

turkey_result (sm.stats.multicomp.pairwise_tukeyhsd): The result object from the Tukey's HSD test.

Returns

pd.DataFrame: A DataFrame containing the results of the Tukey's HSD test.

Example

turkey_result = multicomp_turkeyHSD(factors={'group': ['A', 'B', 'C']}, tag='score', df=data)
turkey_to_table(turkey_result)

multicomp_dunnett(factor:str, exp:List[str], control:str, df:pd.DataFrame, **kwargs) -> scipy.stats.dunnett

Perform a Dunnett's test for multiple comparisons.

Parameters

factor (str): The column name of the factor.
exp (List[str]): The names of the experimental groups.
control (str): The name of the control group.
df (pd.DataFrame): The data frame.
kwargs: Additional arguments for the Dunnett's test.

Returns

scipy.stats.dunnett: The result object from the Dunnett's test.

Example

multicomp_dunnett(factor='group', exp=['A', 'B', 'C'], control='D', df=data)

multicomp_bonferroni(factors:Dict[str, List[str]], tag:str, df:pd.DataFrame, alpha:float = 0.05) -> pd.DataFrame

Perform a Bonferroni correction for multiple comparisons.

Parameters

factors (Dict[str, List[str]]): The factors for data filtering.
tag (str): The column name of the data.
df (pd.DataFrame): The data frame.
alpha (float, optional): The significance level. Defaults to 0.05.

Returns

pd.DataFrame: A DataFrame containing the p-values after Bonferroni correction.

Example

multicomp_bonferroni(factors={'group': ['A', 'B', 'C']}, tag='score', df=data)

Constants

p_value_to_stars(p_value:float) -> str

Convert a p-value to stars indicating the significance level.

Parameters

p_value (float): The p-value to convert to stars.

Returns

str: The string representation of the number of stars. If p >= 0.05, return ''.

Example

p_value_to_stars(0.01)