About the Columns Manager
Columns Manager provides an overarching perspective of an entire column set for quicker decision-making regarding their properties.
Use the Columns Manager to organize columns and to quickly add, edit, and delete column properties.
Access the Columns Manager in one of the following ways:
• Select Cols > Columns Manager.
• Click the Columns Manager button 
in the Columns panel.
• Click the red triangle menu for the columns of a data table and select Columns Manager.
• Right click a column or selection of columns in the Columns panel and select Columns Manager.
• Right click within the Show/Hide Header Graphs area of a data table and select Columns Manager.
Figure 5.5 The Columns Manager Window
The Columns Manager window contains the following information:
Modeling Type
Displays the modeling type of each column which tells JMP how to treat the column’s values during analyses. You can change the modeling type to look at a variable in different ways. See About Modeling Types.
Choose from the following modeling types:
Continuous 
Only numeric data types. Continuous values are treated as continuous measurement values. JMP platforms use the numeric values directly in computations.
Ordinal 
Either numeric or character data types. JMP platforms treat ordinal values as discrete categorical values that have an order. In Graph Builder, if there are more than 50 discrete values, JMP treats the column as continuous.
• If the column has a numeric data type, the values are ordered according to their numeric magnitude.
• If the column has a character data type and a Value Order column property is saved to the column, values are ordered according to the Value Order property. Otherwise, values are ordered according to their alphanumeric data value ordering. However, in special cases where the values have an obvious order, values are automatically ordered appropriately. See Value Order.
Nominal 
(Numeric or Character data types) JMP platforms treat all values as discrete values with no implicit order.
Multiple Response 
(Character data types only) Distinct entries in a single cell must be separated by commas. JMP platforms that support multiple response columns treat each entry in the comma-separated list as a separate data value. If your entries are separated by a character other than a comma, use the Multiple Response column property instead. See Multiple Response.
Unstructured Text 
(Character data types only) These values are generally unique and are therefore not appropriate for categorical analysis. The Text Explorer platform is ideal for analyzing unstructured text values.
Vector 
(Expression data types only) The entries in the cells are column or row vectors. JMP platforms that support the Vector modeling type recognize the vectors and treat them appropriately in calculations.
None 
(Any data type) Use None as a modeling type when a column is not well represented by the other modeling types. For example, a column of pictures or ID values might be assigned the None modeling type. JMP launch windows do not allow you to assign roles to columns with the None modeling type.
Tip: You must use the Columns Manager or the Column Info window to change a column’s modeling type to Multiple Response, Unstructured Text, or Vector.
Name
Displays each column’s name.
Attributes
The Attributes grouping displays the modeling type, subtype, format, and row state attributes.
Type
Displays the modeling type of each column. See About Modeling Types.
Choose from the following modeling types:
Numeric 
The columns contain only numbers (with or without a decimal point).
Character 
The columns contain any characters, including numbers. In character columns, numbers are seen as characters and are treated as discrete values instead of continuous values. The maximum field width for character values is 32,766 bytes.
Row State 
The columns contain row state information, which indicates whether the rows are excluded, hidden, labeled, colored, or marked. See Store Information in Row State Columns.
Expression 
The columns contain JSL expressions that can range from simple entries to complicated statement sequences. In particular, an expression can correspond to a picture, a matrix, or an expression. You can drag-and-drop expressions, such as an image, from your desktop to cells in an expression column. See Expression Role.
Notes:
– If you change a column’s Data Type from Character to Numeric, any character values become missing data values and are not recoverable.
Subtype 
Displays the subtype of each column.
Numeric columns can be saved in a short-integer format to save disk space and increase performance. Numeric columns can be saved as either 8 byte floats, 1 bit integers, 2 bit integers, or 4 bit integers. See The Short-Integer Format.
Character columns may be saved with the compact subtype to save disk space and increase performance. Character columns saved with the compact subtype will display the compact symbol 
. Unless changed in preferences, character columns will be saved by default with the normal subtype and will display nothing in this field.
See Compact Character Columns.
Format
(Numeric columns only) Displays the numeric format of the column. See Numeric Formats.
Flags
Displays the row state flag of each column. See Store Information in Row State Columns.
Expand the grouping to organize columns by the available row state options. Row state categories available to view and sort by include:
Locked 
Columns with this icon have been locked.
Excluded 
Columns with this icon have been excluded from data analysis.
Hidden 
Columns with this icon have been hidden from the data grid.
Scroll Locked 
Columns with this icon remain locked in place while scrolling horizontally across the data table.
Used for Marker 
The values of a column with this icon are in use as the markers in a graph.
Labeled 
Columns with this icon are in use as labels.
Properties
Displays the properties assigned to each column. This field is hidden by default if no properties have been assigned. See Column Properties in JMP.
Summary Statistics
Displays the basic summary statistics including Min, Max, N Missing, N Unique, Distribution, and All Integers. Select Extended Statistics in the Columns Manager window to add Mean, Std Dev, Q1, and Q3. See “Explanation of Summary Statistics”.