User Guide#

The library covers tools used in the following topics.

Defining a problem#

Section: Process Summary and Capability

Shows how to describe a process using statistics calculated from samples of process input and output. Also, provides tools to quickly calculate and plot process capability, a measure of the precision and accuracy of a process.

Measurement and analysis#

Section: Data Analysis

Covers basic inference, regression and ANOVA, and measurement system analysis. The Inference section relates to confidence intervals, hypothesis tests and sample-size calculations for single samples and comparisons between samples. The Regression and ANOVA section further covers statistical inference with linear models that include categorical and quantitative variables, and systems with multiple inputs. The Measurement System Analysis (MSA) section describes an application of ANOVA that quantifies how much variance is contributed by gauges and operators to the total variance in measurements from a product sample.

Making improvements#

Section: Design of Experiments

Discusses how to collect the right data in controlled experiments to answer particular questions using regression and ANOVA. This is important because regression and ANOVA cannot identify the structure of a physical system unless the structure is reflected in the data.

Controlling process#

Section: Statistical Process Control

Shows how to monitor a stable process to observe deviations before they result in defects.

Plotting and notebook tools#

Section: Plots

MQR has a module dedicated to plots. The submodules of the plotting module correspond to other modules in the library. This section describes how the plotting code is structured and the tools MQR provides for displaying plots in notebooks.

Imports#

The pages in this user guide include example code. To ensure that the examples are up to date, they are automatically executed and their output is shown in a cell below the code. All pages that include code make the following imports before any code is run. (The plotting pages make a few further imports; see Additional imports at the top of Plots.)

import numpy as np
import pandas as pd
import scipy.stats as st

import mqr
from mqr.plot import Figure
from mqr.nbtools import hstack, vstack, grab_figure