Plotting¶

The most useful features of manufacturing come in the form of the highly information-dense plotting tools.

Ppk Plot¶

The manufacturing.ppk_plot() takes all of the data - not just a sample - and determines the process capability. This is not a snapshot in time but a look at the entire history. Careful, which this can be can be deceiving!

The :meth:manufacturing.ppk_plot will estimate the distribution based on the input data, calculate the Ppk, mean, standard deviation, and the estimated % out of control for each parameter. The function will also generate a warning if the data appears to be non-normally distributed.

import manufacturing as mn

# the 'data' variable contains a list of integers, floats,
# numpy array, or pandas Series
mn.ppk_plot(data, upper_specification_limit=3.3, lower_specification_limit=3.1)

If manufacturing is used in a jupyter notebook or similar environment, then the plot will display automatically. Optionally, you can pass a matplotlib.figure.Figure instance in order to more directly manipulate the underlying matplotlib.figure.Figure.

import manufacturing as mn
import matplotlib.pyplot as plt

fig, ax = plt.subplots()
mn.ppk_plot(data,
            upper_specification_limit=3.3,
            lower_specification_limit=3.1,
            figure=fig)

ax.set_xlim(3.0, 3.5)  # manipulate the axis as desired

Cpk Chart¶

The manufacturing.cpk_plot() gives snapshots of process capability. In most cases, the average Cpk value should be close to the overall Ppk value. If not, then the process may not be in control.

import manufacturing as mn

# the 'data' variable contains a list of integers, floats,
# numpy array, or pandas Series
mn.cpk_plot(data,
            upper_specification_limit=7.4,
            lower_specification_limit=-7.4,
            subgroup_size=10)

Zone Control Chart¶

Perhaps the most useful chart is the manufacturing.control_plot(), also known as a Zone Control Plot. This plot will highlight up to 8 different rules or violations based on the input data set. If a control chart rule is not violated, then it will not be placed on the chart.

There are three different types of control charts defined within manufacturing:

manufacturing.x_mr_chart() for small data sets

manufacturing.xbar_r_chart() for subgroups between 2 and 10, inclusive

manufacturing.xbar_s_chart() for subgroups larger than 11, inclusive

Using the manufacturing.control_plot() function will automatically select the appropriate control chart type based on the number of data points supplied.

Control Chart Rules by Severity :header-rows: 1¶
Violation	Pattern
beyond limits	Point is beyond the limits
zone a	2 out of 3 consecutive points in zone a or beyond
zone b	4 out of 5 consecutive points in zone b or beyond
zone c	7 or more consecutive points on one side of the average (in zone c or beyond)
trend	7 consecutive points trending up or trending down
mixture	8 consecutive points with no points in zone c
stratification	15 consecutive points in zone c
over-control	14 consecutive points alternating up and down

import manufacturing as mn

# the 'data' variable contains a list of integers, floats,
# numpy array, or pandas Series
mn.control_plot(data)

Depending on the data set, the above command could result in the creation of an \(X-mR\) chart, \(\bar{X}-R\) chart, or \(\bar{X}-S\) chart.

Anatomy of a Control Chart¶

A control chart is an information-dense representation of data coming off of testers. The manufacturing control chart has several sections that warrant further explanation:

chart type - the chart type, usually based on recommended groupings

group size - when present, indicates how many samples were grouped to present each datapoint

distribution histogram - the statistical distribution of the data

more data indicator - when present, indicates that data was truncated before display in order to to not present so much information that the plot becomes unreadable

out of control indicators - when present, indicate that the process is out of control or nearly so

upper control limit - the upper control limit as calculated from the data on the plot

avg value - the average value as calculated from the data on the plot

lower control limit - the lower control limit as calculated from the data on the plot