mqr.process.

Sample#

class mqr.process.Sample(data, conf=0.95, ddof=1, name=None, num_display_fmt='#.5g')#

Data and descriptive statistics for a single sample from a process.

Construct using a pandas Series (ie. a column from a dataframe). Intended for use by a Study object.

Attributes:

namestr: Name of the KPI or measurement.
conffloat: Confidence level to use in confidence intervals.
datapd.Series: Sample measurements.
ad_statfloat: Anderson-Darling normality test statistic.
ad_pvaluefloat: p-value associated with ad_stat.
ks_statfloat: Kolmogorov-Smirnov goodness of fit (with normal) test statistic.
ks_pvaluefloat: p-value associated with ks_stat.
nobsint: Number of measurements in the sample.
meanfloat: Sample mean.
semfloat: Standard error of mean.
stdfloat: Sample standard deviation.
varfloat: Sample variance.
skewnessfloat: Skewness.
kurtosisfloat: Kurtosis.
minimumfloat: Smallest observation.
quartile1float: 25th percentile observation.
medianfloat: Median observation.
quartile3float: 75th percentile obsevation.
maximumfloat: Largest observation.
iqrfloat: Inter-quartile range.
conf_meanConfidenceInterval: Conf interval on the mean.
conf_varConfidenceInterval: Conf interval on the variance.
conf_quartile1ConfidenceInterval: Conf interval on the 25th percentile.
conf_medianConfidenceInterval: Conf interval on the median.
conf_quartile3ConfidenceInterval: Conf interval on the 75th percentile.
outliersarray_like: List of points falling further from a quartile than 1.5 * iqr.

Examples

In a jupyter notebook, sample summaries are shown as HTML tables:

>>> data = pd.read_csv(mqr.sample_data('study-random-5x5.csv'))
>>> mqr.process.Sample(data['KPI1'])

produces