# Distribution Analysis

The Distribution Analysis report includes several tools that can be used to gather advanced information regarding a part's features and evaluate their capability.

You can use the *Filters *in the sidebar on the left side of the page to select a Part Number and Feature belonging to your company or one of your suppliers, which will populate the page's available charts. You can also filter by Date Range, Run, Revision, Machine, or Operator to drill down to specific measurements on the selected feature.

Each chart provides the following information about the charted data.**N: **Number of plotted measurements in the sample**Average**: Average measurement value in the sample**AD**: Anderson–Darling statistical value

The parameters used to describe the probability density function of each distribution type are also shown on each plot.**Location**: Amount of transformation on the x-axis of a probability density function**Scale**: Amount of compression/stretching of the probability density function**Shape**: Representation of the shape of the probability density function

__Probability Plots__

The Probability Plots report can be used to evaluate the type of distribution that best fits a dataset. This is particularly useful for finding out whether data fits a *normal distribution*, which in turn determines the applicability of process and capability indices such as PPK and CPK, and how these values should be calculated.

After you have selected and filtered your data, Net-Inspect will transform each measurement to fit a series of charts which plot each measurement according to the proportion of measurements that have a lesser value. Each of these charts then plots the set of measurements against one of the following distribution types:

- Normal
- Log Normal
- Exponential
- Logistic
- Weibull
- Largest extreme
- Smallest extreme

Net-Inspect performs an Anderson–Darling test for each distribution type and derives a P-value, which represents the likelihood that the measurement data conforms with that distribution; the larger the P-value, the more likely the fit. The probability plots are ordered by P-Value from left to right, with the distribution in the top-left of the page representing the "Best Fit" for the data.

__Histograms__

Histograms provide a visualization of the ranges that a defined set of measurements occupies. The more measurements that fit into a particular range, the taller the bar representing that range will be on the histogram.

Each histogram is compared against the "curve" associated with and predicted by the distribution types listed above; the closer the fit, the more likely it is that the data conforms with that distribution type. Data that fits a "normal" distribution will take the shape of a bell curve, while data that forms other shapes may conform more closely with other distribution types.

Histograms are ordered by P-Value from left to right, with the histogram and distribution curve in the top-left of the page representing the "Best Fit" for the data.

__Control Charts__

The Control Charts report provides both a chart of individual measurements, as well as a "moving range" chart showing the change between two consecutive data points. A histogram is also provided with each chart to provide additional information.

__Add to Favorites__

To add a custom chart to your dashboard, please see Adding to Favorites.