statsmodels.stats.oneway.AnovaResult#

class statsmodels.stats.oneway.AnovaResult(*, statistic, pvalue, df, df_num, df_denom, nobs_total, n_groups, means, nobs, vars_, use_var, welch_correction, df2=None, df_num2=None, pvalue2=None)[source]#

Result of anova_generic and anova_oneway.

Parameters:

statisticfloat: Test statistic for k-sample mean comparison which is approximately F-distributed.
pvaluefloat: If use_var="bf", then the p-value is based on corrected degrees of freedom following Mehrotra 1997.
dftuple: Degrees of freedom (df_num, df_denom) for the F-distribution used for pvalue.
df_numfloat: Numerator degrees of freedom.
df_denomfloat: Denominator degrees of freedom used for pvalue.
nobs_totalfloat: Total number of observations across all samples.
n_groupsint: Number of samples being compared.
meansndarray: Mean of each sample.
nobsndarray: Number of observations in each sample.
vars_ndarray: Residual (within) variance of each sample.
use_var{“unequal”, “equal”, “bf”}: The use_var option that was used to compute the test.
welch_correctionbool: Whether the Welch correction was included in the test statistic.
df2tuple or None: Degrees of freedom (df_num2, df_denom) for the Brown-Forsythe 1974 p-value. Only set if use_var="bf", otherwise None.
df_num2float or None: Numerator degrees of freedom for the Brown-Forsythe 1974 p-value. Only set if use_var="bf", otherwise None.
pvalue2float or None: p-value based on degrees of freedom as in Brown-Forsythe 1974. Only set if use_var="bf", otherwise None.

Attributes:

df
df2
df_denom
df_num
df_num2
means
n_groups
nobs
nobs_total
pvalue
pvalue2
statistic
use_var
vars_
welch_correction

Notes

Unpacks as statistic, pvalue = result. Other values are only accessible using attributes.

Methods

Properties

`statistic`
`pvalue`
`df`
`df_num`
`df_denom`
`nobs_total`
`n_groups`
`means`
`nobs`
`vars_`
`use_var`
`welch_correction`
`df2`
`df_num2`
`pvalue2`