Data Preparation
Preparing your data for troubleshooting the problems experienced on your plant is the first step in the 7 step Rapid Process Troubleshooting methodology.
The Data Preparation step is accessed from the Troubleshooter's project bar as step 1.
In this step, data is imported from historical sources, a number of mathematical and statistical operations can be applied, and data can be manipulated. This enables you to create a dataset that is relevant and appropriate for troubleshooting your process. This dataset is created in preparation of modeling the system to analyze how to reduce process variation.
Data requirements and limitations
During data preparation, you are limited by the number of columns of the dataset, but not the length of the dataset.
Read more on dataset limitations
Procedure to create a data recipe
A data recipe is a set of instructions and actions to perform on your data before loading it for visualization in step 2 of the Troubleshooter.
From the list of operations on the palette, select how to manipulate your data to create a representative dataset for your process. Click the [+] next to each category, in order to see the list of available operations. Different operations have different effects on the datasets, either affecting the current dataset, or creating a new dataset. These functions are explained in the individual help documentation for each operation.
Note: Holding your cursor over the operation name will show a tooltip briefly describing the operation function.
The following categories of operations can be used to create a data recipe:
-
Import a dataset.
-
Add or remove fields as required.
-
Modify the dataset by replacing certain values, manipulating time and scale properties or renaming fields.
-
Transform datasets by changing dataset formats, converting data types, resampling, correlating and transposing data.
-
Filter datasets on values or timestamps, or by using scripting expressions.
-
Combine datasets in a number of different ways.
-
Export datasets either to a file for analysis, or to another application for further manipulation.
Related topics: