JSL language elements

JMP's comparison operators and logical operators are short-circuit operators, that is, they will stop working as soon as the result is returned.

It should be noted: ① When the latter parameter depends on the previous parameter. ② Overly cumbersome calculations lead to calculation errors in JMP. ③When depending on the last parameter to be executed. Be careful about placing assignment statements so that they are not placed inside expressions that never execute.

Date/time data is handled internally as seconds since midnight on January 1, 1904.

Use Date DMY(1,12,2003),today()

For example, day, year, day of week, week of year, time of day, etc.

Perform common arithmetic operations on date/time data.

The operators in minutes, in hours, in days, in weeks, and in years are used to re-express time intervals using units other than seconds.

The glue or ; semicolon operator evaluates each argument consecutively and returns the result of the last argument. The semicolon is a terminator in other languages, but in JSL it is only a binary operator used to combine expressions. The result returned is the rightmost expression result.

First function first is similar to glue function, except that it returns the first parameter instead of the last parameter.

For(init,while,iter,body), the function for will calculate the init expression, but only once. The while expression is then tested. If not true, it breaks. As long as while is true, the body and iter expressions are executed.

while(condition,body), while is a function related to for, it can repeatedly test the condition, and as long as the condition is true, its main script will be executed.

summation(initial argument, limit argument, body expression) returns the sum of the calculated values ​​of the body expression, each time increasing the variable from the initial argument until it is greater than or equal to the limit argument.

Product is similar to summation, except that it multiplies the main results instead of adding them.

if (condition 1, result 1, condition 2, result 2..., else result), calculate the conditions in order until the first non-zero result is returned, and return the corresponding result statement, otherwise will return elsesult results.

y = Match(x, v1, expr1, v2, expr2, ..., exprElse) finds the first vN parameter equal to x and calculates and returns the corresponding exprN parameter. If there is no value equal to x, calculate and return exprElse parameter.

y=choose(i, expr1, expr2,..., exprElse), calculates and returns the i-th expr parameter. If there is no i-th expr parameter, calculates and returns the exprElse parameter.

interpolate(x, x1, y1, x2, y2, ....) or interpolate(x, xmat, ymat) will linearly interpolate the y value corresponding to the given x value between two points.

step(x, x1, y1, x2, y2,...), calculates the y corresponding to a given x through step function fitting instead of linear fitting between points. It will return the maximum value less than or equal to x. The yi parameter corresponding to xi.

throw() function, the script will stop running on its own. If you want to skip an error-prone part of your script, you can include it in a Try() expression. try(expr, "catchexpr"), when the first expression raises the expression by evaluating throw, it will immediately stop evaluating the first expression and return no result, and evaluate the second expression.

Stored expressions can be used as macro functions. Use expr to store expressions to delay evaluation until the global variable is actually called

Function({x},expr) specifies local variables in {} and declares expressions directly.

Local symbols: Variables can be defined as local to a function so that they do not affect the global symbol space. Such functions need to place the values ​​of local variables separately at each level of function call calculation.

Recursive recurse(x1) calls the defined function recursively until the recursive condition is reached.

The Local function allows you to handle local variables without affecting the surrounding namespace, for example, the called script has a variable with the same name as the global variable. The general form is: Local({list_of_variables_or_assignments}, body ); all names in body will be parsed locally.

The include function is used to open a script file, parse the script, and execute the script. Generally means include ("pathname" <<<parse only>)

Loading and saving text files load text file ("PATH") and save text file ("Path", text) can be used to process text files in JSL.

XML parsing operation---unknown on time,

Dynamic link library DLL file, JSL has three commands to load the DLL and then call the functions exported by the DLL.

main menu ("Command" <window name>) Note: Use English, not Chinese, ah ah ah ah

{a,b,c}={1,2,3}; {a,b,c} =10, {{a},{b,c}}, abc increases by 1 -- decreases by 1

If the list is an assignment list or a function list, you can also use quoted names as subscripts: X={a(1),b(3),c(5)}; show(X["a"]);

For example dist=expr(distribution(column(height)))

If you execute a script, you only need to use dist;

Only at the outermost level of the formula can eval work.

nameexpr looks for the expression stored in its argument. But it does not calculate, and only unwraps one level to get the expression.

Parsing only involves scanning strings into language expressions according to grammar. Suppose I have read a valid JSL expression into a string and now want to evaluate it, the function parse will return the expression. To calculate it, use the Eval function.

Otherwise, use char to convert the JSL expression into a string. For example: y=char(expr("a=1")). In addition, char also allows setting field width and decimal place parameters.

To convert a string into a numeric value, you need to use num

Positional operators: refer to operators that can act directly on lists or expressions. They all have from or into in their names. The first parameter of a positional operator must be an L value, an entity (such as a global variable) whose value is set.

Non-positional operators: Must declare the list directly or refer to a name that can be used to calculate the list. The name does not contain from or into. The processed list or expression will be returned without changing the original list given in the first argument.

Substitute: Both substitute and substitute into can be used to find all items that match a pattern in a list and replace them with another expression. These parameters need to be quoted using expr.

Operators for lists

Confusion warning! The current row of the script is the same as the row or data table to be selected (highlighted) in the data table The current cursor position in the window is not relevant

1. If the name is followed by a pair of parentheses (), it can be regarded as a function 2. Then treat it as a local variable 3. If the prefix is ​​not:, it will be regarded as a global variable 4. If the prefix is ​​not::, it will be regarded as a data table column or table variable. 5. Then, treat it as the platform launch name (e.g. distribution, binary) 6. Then, if it is used as an assignment target (as an L value), a global variable is created and used.

When an unscoped name is reparsed, it refers to a column in the data table rather than a global variable.

If the data table has a table variable with that name, that table variable takes precedence

1. Prefix colon: indicates that the name can only refer to table columns or table variables, but not global variables. Prefix: refers to the current data table context.

2. Infix colon: Specify which data table contains the column by extending this notation with a data table reference. Equivalent function as column(data table ref, name), e.g. currentdatatable():name

The global scope operator can be used to force names to be resolved into global variables. The prefix double colon: indicates that the name can only refer to global variables, not table columns. , equivalent function as global

NOTE: Each new Caption will hide the previous one, i.e. only one Caption window will be available at a time. To close the caption without displaying the new one, use the specified parameter Remove,caption(remove).

pick file() locates the path information of the file and saves it as a string

The pick directory() command has an optional string displayed as a prompt message

files in directory(path), get the list of file names in the specified directory

dialog and column dialog, dialog creates a new dialog box, and column dialog allows the input of column parameters and associated variables of the data table

arrangement

Assign dialog to a variable. JMP will store the list returned by the dialog function in the variable. If the variable is called, the list will be called instead of calling Dialog.

Unload results from the dialog box. Use removefrom to eliminate unnecessary results.

Open commonly uses path parameters. If there are no parameters, a dialog box will be displayed to prompt the user to locate a file.

complex

open database Use ODBC to open the database and extract data to JMP data table

1. dt=newtable("my.jmp")

2. Create a new data table to avoid windows and use the invisible keyword. For example: dt = newtable("abc",invisible,newcolumn("x"),addrows(10))

Reference an open table

Use the << operator to send a message to an object, which is also the Send function

A series of messages can be stacked in one statement

After the run script command is sent to the data table, it will first find the specified property (that is, the script) and then run it as a JSL script.

Use the close command with a data table reference as a parameter.

To revert to a recently saved data table, send a Revert message to the data table. Before restoring, you will be asked whether to save the current data table.

NTable() can return the number of open data tables

The New Table and Open commands are used to create objects that do not exist. Once created, these objects know what to do with them, so you can send them messages requesting changes. To close such an object, a command must be used on its container, since objects cannot delete themselves. Close is such a command.

Use current data table(name) to make this data table the current data table.

The current row is determined by the value of the pseudo variable of row(). You can assign a value to row(), or you can use the iteration command for each row()

Get a list of selected columns

To select certain rows, use select rows

To set the column name of the cell in the current row, please use an assignment statement.

If you want to get the value, just specify the name,

If setting a specific row of a column, rather than the current row, use the row number as a subscript.

Confusion warning! Note that the subscript must be a row that exists in the table. The default line number is 0, so something like :name Languages ​​that reference line 0 in this way will generate an invalid line number error.

By default, the current line number is 0. The first row in the table is row 1, so row 0 actually exists

row(): operator gets or sets the current row number.

Nrow and Ncol operators are used to count the number of rows and columns in a data table.

Name, save and print

Log and layout

Create and delete columns

Column reordering

To iterate a script over each row of the current data table.

Lag returns a value listed n rows before the current row

Dif returns the difference between the value in the current row and the value in the previous n rows

See learning script for details

This command mainly collects summary statistics of the data table and stores them in global variables, while summary collects summary statistics and displays them in a new data table.

Definition: Metadata includes information such as data sources, comments about each variable, scripts used to process the data, column analysis roles, etc.

table variable

Properties/Scripts

Delete variables, properties/formulas

on open nature

Script to start on startup

script

Control formula evaluation

Set value without using formula

Control display updates

Resize data table

Convert between matrix and data table

Row messages point to data table references, and most only affect the currently selected row. Many line messages are generally of no use in scripts unless you are writing a script that simulates the interactive use of JMP.

Add row

Delete row

Select row

Commands in the table menu:

column name(n) returns the name of column N

get column names retrieves a list of the names of all columns in the data table

Get selected columns

Move columns

data column object

Access cell value via column reference

Send message to column

You can use messages that correspond to data table columns to control various properties or characteristics of the column, including its name, data, and metadata.

For example, hide, exclude, label, and scroll lock can be activated through scripts. 1 means turning on the column attribute, and 0 means turning it off.

name

value

value label

Data and modeling types

display format

Preselect roles

formula

Range checking and list checking

Set, retrieve, and remove column properties

locking

script

1. Manual--

2. Numerical method: Set the row status using an expression whose calculation result is the row status value.

Most row states (exclude, hide, mark, select) are boolean, that is, they are on or off, that is, 1 or 0. Marking includes 0~15, and color options are 0~84, where 0~15 is basic Color, 16~31 is for darkening of the same color, 32~47 is for brightening, 48~63 is for darker, 64~79 is for highlighting, and 80~84 is for gray.

The row status column is neither numeric nor character, just row status. And it only stores the row status information, but does not make the information effective.

In JSL, you can use the Row state operator to directly get or set the row state. row state(n) can get or set the status of the nth row. If no arguments are given, sets or gets the current row.

To set the current state of a row, please use row state() on the left side of the assignment

To copy several, but not all, current row statuses to the row status column, use a script like the following

Multiple state settings are combined by combine states, and the final row state is feature accumulation.

Get the row status of a row

Row status columns can also have multiple characteristic row statuses as their values

color state sets color row state

combine states merge features

excluded state exclude row status,

hidden state hidden row state

hue state returns the row state value with the hue state portion set to the specified value. Need to be used in conjunction with shade state to display effective colors

Shade state returns the row state value with the color light and dark state part set to the specified value. It needs to be used in conjunction with hue state.

labeled state label row state, label of data row

marker state Mark row state, for example, the point becomes ×

selected state selected row state

A row state knows not one feature, but many features.

color of returns the color component of the specified row status. If it is an L value, it will change the color of the current row of the current data table.

Also includes excluded, hidden, labeled, marker of, selected

Place the row status expression to the right of the assignment

Access only a certain part of the row state

To access row status, place the row status expression to the left of the assignment (the L value)

Set row status

Use combine states to gather settings for various states together

Create row status directly in column

Validate row status in column

This section describes Boolean states, i.e. switching conditions

excluded, hidden, labeled and selected set selected indicators

Excluded state, hidden state, labeled state and selected state perform opposite operations Gets a row status value or sets a row status condition to true or false based on a parameter.

color of returns or sets the color indicator

marker of returns or sets the marker indicator

Color State and Marker State are similar to Color Of and Marker Of, but in opposite directions. The -Of function converts the actual state into an indicator, while the -State function converts the indicator into a state.

If using RGB values, there should be a list for each color, with percentages in red, green, and blue order.

The hue state, along with the shade state, is an alternative to the color state for picking colors.

Hue and Shade do not have the "-Of" operator. Color Of returns the equivalent Color State pointer to the color row state. icon where the state has been set using Hue State and/or Shade State.

The As Row State operator assigns row status through internal numeric codes. This operator simply converts an integer Convert to its equivalent row state

JMP has col maximum, col mean, col minimum, col N missing, col number, col quantile, CV (coefficient of variation), col standardize, col std dev, col sum, maximum, mean, minimum, Nmissing, number, std dev and sum function, these are precomputed functions

Note that precomputed functions ignore excluded row status, so the calculation will include any excluded rows. To follow summary statistics of row exclusions, you can use distribution platforms.

Functions whose names begin with col all work column-wise, or down through the values ​​in a specified column, and return a single number.

Functions listed without "col" work row by row across the values ​​of the specified variable and return column results.

JMP can edit formulas in a structural manner. There is no difference between formula columns created through the calculator window and formulas created through JSL using commands like New column(..., formula()) or col<<formula()

The first and simplest method is to start the platform from a script. The script will assign the object reference to the global variable.

The second method is to assign the value by using a subscript to reference the platform name. This method is suitable for started platforms.

1. You can ignore the parameters of the Boolean option to switch the state

2. If the menu provides several options separated by commas or slashes, such as means/anova/T test, you can use any command. If several commands have the same alias, the first command takes precedence over the scripting language.

3. If the display is changed, such as adjusting the graphic size, it will not be reflected in the saved script. The saved script only records options related to analysis and does not record demonstration details.

4. If the submenu item represents a command rather than a setting, the corresponding script is the menu item itself without a parent. For example, One-Way Analysis has the menu item Non-parametric and has three commands in the sub-menu, including the Wilcoxon test. In the script, only the sub-item name will be used.

5. If the submenu item is a set value rather than an independent command, provide the parent item in the script and the submenu option as its parameter.

6. The script returned from the platform usually looks different from the script written by yourself. For example, you can use a simple script to start the function, distribution

To send several messages, you can add more << operators or more Send parameters

Because << is an ignore operator, it merges parameters and works differently than when combining parameters. You can use additional << symbols to stack multiple messages and execute them all in order (from left to right).

You can use grouping parentheses to send a message to the result of sending the message. In this case, the associated grouping has no order.

Another way to stack messages is to send a list of messages

1. Try the program through the interactive interface, then study the saved scripts.

2. Research the interface in the platform window, and the menu items of the pop-up menu and the status menu have equivalent JSL with the same name and parameters.

3. Go to the "Index" tab in the "JMP Start Page" window, click the "Object Button", find the required object type, and click the corresponding list item.

4. Show properties (object ref) lists all messages that the object can receive in the "Log" window

[Sublist] refers to a set of commands placed in a submenu.

[Boolean value] is used to close and open options. Their parameters are usually 1 or 0. If no parameters are specified, the opposite state will be switched.

Actions and New Entity are general-purpose commands that typically result in a dialog box in the user interface being displayed. There are no specific standards for the parameters of the operation. You can try the project in the interface first, and then study the scripts saved by the platform.

[Operation Selection] and [Enums] parameters have specific sets of options.

If you want to use an expression that will be evaluated for the column parameters, put the name inside the Eval or evallist function

The column parameters of the platform startup script can also be a list with curly brackets { }, so the following scripts are valid

After sending the command Action to the platform, it just evaluates the expression, no matter what it is. This command allows continuous invocation of the platform.

Platform startup allows the use of the invisible option, which suppresses the display of windows. Using this option in a model fitting script suppresses the model dialog box and results window.

The invisible option also applies to "Table" menu operations, suppressing window creation (not just hiding). But after using the data table for the first time, it automatically deletes itself when closing the analysis that uses it.

Add the title command to the startup request to specify the title.

Among them, J(NR,NC,VALUE) creates a matrix of the same elements.

The reference of the platform report is from the display box >> report subtitle >> row number

When saving a script based on a By group, a where clause is added showing which group is used.

Scripts for the entire analysis window, which are linked together

If you submit this script, the same analysis as the By group script will be regenerated, except that the results for each group will be displayed in a separate window. You can use new window and Vlist box to glue the results together

The where clause is applicable to all platforms that support By group processing, and any JSL expression can be its parameter

Some platforms have substructures that also have objects inside them that are addressable via the internal Send operator.

The save script command works on almost all platforms

The report command can access the display surface and control appearance information.

By parameter is supported on almost all platforms

The Title command is available in all platforms and will replace the title

Many platforms mention that messages can be sent to axes within the platform, and the value axis

The first, basic startup, applies the same options to many variables

The second, detailed startup, each column can be analyzed using different options

quantiles, quantiles, Boolean parameters

moments, summary statistics, boolean parameters

more moments more summary statistics, Boolean parameters

horizontal layout horizontal layout, Boolean parameter

histogram displays histogram, boolean parameter

std error bars histogram option, display error bars, Boolean parameters

count axis histogram option, display count axis, boolean parameter

prob axis, histogram options, display probability axis, Boolean parameters

density axis histogram option, display density axis, Boolean parameter

show percents histogram option, display percentages, boolean parameter

show counts histogram options, show counts, boolean parameters

CDF plot Display CDF plot Boolean parameters

test mean (number), mean test, non-boolean parameter

test std dev (number), standard deviation test, non-boolean parameter

confidence interval (probability), specifies the confidence interval, non-boolean parameter

prediction interval (confidence level, n future observations) prediction interval. You can use Boolean parameters to open the dialog box, or enter startup parameters to start directly.

tolerance interval Tolerance interval, you can use Boolean parameters, then open the dialog box. Or enter startup parameters to start directly.

capability analysis (USL (number), LSL (number), Target (number)) Capability analysis, enter specification limit parameters

fit distribution (option) fits the specified distribution, all fits all distributions and compares

ppk capability labeling In capability analysis, switch the label to pp instead of cp Boolean parameter

save (option) save, create a new column in the data table to save

axis settings

histogram color (number) histogram color

Too many, more complicated

The fitted curve can be addressed by subscript curve notation

Messages for sending sub-objects can also be sent directly in the startup statement in the form of a list, after any parameters required by the sub-command.

If you use the group by function to generate curves for the groups identified by the grouping column, the generated script will use a fit where clause to define a subset of each curve. Plotted on the same graph. BY is drawn as two pictures.

Options for fitting functions

Options for various types of curves

Optional options

Show options

XAxis and YAxis messages can be used to direct Axis commands to the drawing.

Shows how the distribution of categorical responses differs across groups

Options include

A logistic fit shows how the distribution of categorical responses differs as a function of continuous variables

Options

Analysis shows how paired variable pairs differ in their means

Options include

Model Fitting Dialog Box

Nonlinear fitting, by least squares or a custom loss function fitting a model defined by a formula with parameters to be estimated

Multivariate Analysis: Display Scatter Plot Matrix

Recursively separate data based on the relationship between Y and x, creating a decision tree

Discriminant analysis, classifying multiple y values ​​into various groups of x values

Options

Partial least squares, used to predict y for multiple x, especially if there are more x than the number of rows

neural network fitting

Options

hierarchical cluster analysis

The Get Column Names command returns column names as a list of character values. It is useful when doing two-factor clustering, and the names are in the same order as when column clustering.

k-means cluster analysis for large data sets

Build a time series graph

Plot data for a numeric column or plot statistics

Platform options

Options for each Y

Options for each level

Send the message to the subscripted Y or Level, set the properties of a single y or a single level

Send messages to the X-axis, overlaid Y-axis, or single Y-axis. Here, X Axis refers to the common category axis (defined by shared by all plot cells), Y Axis[i] refers to the i-th Y axis, Y Axis without a subscript applies the command to all Y axes. ch = Chart(X(age), Y(Min(height), Mean(height), Max(height))); ch << (x axis << { axis name("abc"), No Major Tick Mark(0) }); ch << (overlay y axis << show major ticks(0)); ch << (y axis[2] << show major ticks(0));

Overlay several rows or markers on the y-axis based on shared variables on the x-axis

Platform options

To change a specific axis, use the Send operator with Left Axis or Right Axis. To change the selection of a specific variable item, use the Send operator with the name of the Y column.

3D scatter plot

Contour map

Options

Example

Add a legend interactively using the row legend command

The graphics script is set in the gragh box command in the new window command

Plot function

Get the properties of the graphics box

Draw lines, arrows, points, and shapes

Add text

color

transparency

Fill mode

line style

Use pixel controls

The handle and mousetrap functions are used to make interactive graphics that respond to clicks and drags. Another method is to use a button box, slider box, or global box to place buttons or slider controls outside the graphic.

handle, create an interactive graphic that responds to dragging.

Create interactive graphics that respond to mouse clicks

There are 5 drag functions that perform similar functions to handle and mousetrap, but with multiple points at once. 1. drag marker draws n markers 2. Drag line uses n vertices and n-1 line segments to draw a connecting line 3. drag rect draws a filled rectangle using the first two coordinates, ignoring any other coordinates. 4. drag polygon uses n vertices to draw a filled polygon. 5. Drag text draws a text item at the coordinates. If it is a list of text items, draw the i-th list item at the i-th (x, y) coordinate. If the list items are less than the coordinate pair, the remaining points will be repeated. The last item.

Troubleshooting: Interactive graphics are not working as expected, make sure handle or mousetrap coordinates are provided with initial values

picture display type. The picture data type in JSL is used to store pictures of JMP output or formulas.

Using the << operator, you can send messages to the display being constructed without having to send them as messages after they appear. This allows you to distinguish between computed subparameters and option parameters. It also helps make it clearer which parameters are options and which are scripts to be run in the graph.

Parameters used to control settings in the graphics box

Function to construct interactive graphics box

A function that plots a function within a graphics box

Function for drawing in pixel coordinates within a graphics box

Vlistbox glue boxes together vertically

Hlistbox glues boxes together horizontally

Nest Vlistbox and Hlistbox together

tablebox is a special Hlistbox whose elements are string and numeric columns

outlinebox creates an outline hierarchy

picturebox glues axes, boxes and labels together to make graphics

JSL also has some other display box types for customizing the display: buttonbox, slider box, tab box and global box. In addition, there is support for editable text boxes text edit box and for making trees similar to those in the cause-and-effect diagram platform. box hier box

A special JSL value called a display box reference can own or reference a display box (this book uses db as a placeholder for any display box), dt is a data table, and obj is a scriptable object.

Display box references can be created using report messages to facilitate the top of the display tree associated with scriptable platforms

subscript

wildcard

Display references can be used to send scripts to display elements via the send or << operators

close can switch the frame node back and forth between the closed and open states. Can also contain a boolean parameter, representing 0/1

Send window messages to find windows nested under other windows

The send to report and dispatch commands can be used together to customize the report appearance. For example, turn on and off the outline node, adjust the size of the graphics box, or customize the color in the graphics box.

send to report contains a list of commands that can be sent to the display tree.

dispatch is used to send commands to a specific part of the display tree. It has four parameters, the first of which is a list of outline nodes that need to be traversed to find the required display tree part. The second and third parameters work together, the second parameter is the name of the display element, and the third parameter is the type of the display element. The fourth parameter is the command to send

How to get a single number from a table

It is also possible to get the values ​​from the box as a matrix which can then be used for further calculations and written to a data table. You can also create data tables directly.

Adjust axis dimensions

save Picture

Record report log

window finds a window by title and returns a reference to that window.

Subscript used to display the box

You first need to start the new window, start with a title, and then list the items to be constructed in the window. All display box constructors end with "box"

Assemble partial pseudo-tables of matrix JSL using OutlineBox, HListBox and Matrix Box

Can be programmed to avoid iterating the outline box

Sometimes you don’t know how many display boxes need to be displayed in future reports?

append This message adds a display box to the existing display. For example, in a script, construct a single empty box and then use <<Append to add boxes within it for each variable in the analysis.

The prepend message works similarly to append, except it adds the item at the beginning of the display box instead of at the end.

Using the delete method of the display box will make the display box and all its children disappear.

The sid append message can add a display element to an existing tree. Display box or display segment

To update the numerical column in the display box, use the set values ​​command

Usually JMP will automatically wrap the text in the text box. The default wrap point can be overridden with the set wrap(n) message, where n is the number of pixels to be displayed before the wrap point.

You can also use bullet point (1) to use bullet points. This message sent to the text box will place a small bullet point in front of the text and make subsequent lines in the text box indented.

JMP has three special types of display boxes, which are usually not seen in the JMP platform: button box, slider box, and global box. Used to build custom windows with interactive graphics.

The slider box draws a sliding bar control that is used to select any value of the specified global variable, and its range is determined by the specified min and max. When the slider is moved, the value specified by the current position of the slider will be assigned to the global variable, and the graphics will be updated accordingly. Therefore, the slider box is another method for parameterizing graphics.

button box draws a button using the specified name. When the button is clicked, the script will be executed.

The global box displays the name and current value of JSL global variables. The user can assign a new value to this global variable by editing the value directly in the window and pressing Enter or return to commit the changes.

Example: Combine graphics with two sliders and a button,

Modal and non-modal controls

If you give a construct a name, that name becomes a reference to the window, and the window will have the display box inside it. Subscripting lets you send messages to the window's display box.

It may be necessary to construct another display to simply combine the results in the JMP Analytics Platform

Improved this instance by collecting coefficients from the user via a dialog box. This section further develops this example into a complete custom platform, first popping up a dialog box prompting for coefficients, then finding the roots, and finally displaying the results and graphics in a custom window.

It's relatively easy to place objects in the log window, but the harder part is knowing how to handle the log itself

A matrix is ​​a rectangular list of numerical rows and columns. JSL supports matrix values ​​as well as multiple matrix operators. Statistical methods are generally used in matrix expressions. As used in this chapter, scalars refer to simple non-matrix numeric values. If a matrix has only one row or column, it is generally called a row vector or column vector, sometimes just a vector.

If you store multiple numeric expressions in a list, you can also use the matrix function to convert them to a matrix.

The function identity can be used to construct an identity matrix of specified dimensions.

Function J can be used to construct a matrix with the specified number of rows and columns as the first two parameters of the matrix and the remaining elements as the third parameter.

The index function is used to generate a row vector of integers from the first parameter to the last parameter. The double colon: is an infix operator with the same effect:

The function shape can be used to reshape an existing matrix by changing its row distribution to a specified dimension.

Rows and columns can be deleted by the following assignment operation

ncol/nrow can return the number of columns and rows in the matrix or data table respectively.

Addition, subtraction, multiplication, and division operations can be performed on matrices. The standard multiplication operators become matrix multipliers instead of element multipliers.

If given two matrix arguments, the "*" and "/" infix operators and the equivalent multiply and divide functions will perform matrix multiplication and matrix division operations. But if given a matrix and a scalar, they will perform element-wise multiplication and division. You can also use the matrix mult function to specify matrix multiplication To force element-wise multiplication, use: * or the equivalent emult function, scalar multiplication is commutative, while matrix multiplication is not To force element-wise division, use:/ or the equivalent ediv function

Other operators and functions perform operations based on the elements in the matrix. For example, if A is a matrix and B=SQRT(A), the SQRT function is executed on each element in matrix A.

Such functions include

concat can combine two parallel matrices into a larger matrix, the number of rows must be the same. Equivalent infix operator ||

vconcat can stack two connected matrices together, and the number of columns must be the same. Equivalent infix operator |/

Both concat and vconcat can combine empty matrices, scalars and lists

As an alternative to the coalescing operators ||= and |/=, a||=b is equivalent to a=a||b and a|/b is equivalent to a=a|/b.

diag can be used to create a diagonal matrix from a square matrix or vector

For example, D=[1 -1 1], Diag(D)=[1 0 0, 0 -1 0, 0 0 1]

vecdiag creates vectors from the diagonal elements of a matrix

The trace function will return the sum of the diagonal elements of the square matrix, which is the trace of the matrix:

The transpose function is used to transpose a matrix. The back quotes are its equivalent postfix operators, namely A' and At

Matrix data tables can be converted to each other

The expression A[I,J] means to extract the element of row i and column j and return a scalar number.

You can also specify a list or matrix of subscripted values, and the result is a matrix of selected rows and columns

Single subscripting treats a matrix in the same way as converting all rows into a single column, with the columns connected end to end. Double subscript A[i,j] produces the same result as single subscript A[(i-1)*ncol(A) j]

Use the subscript parameter 0 to select all rows and columns

New values ​​can be inserted into the matrix via subscripts. The subscript can be a single subscript, a subscript matrix or list, or a 0 subscript for all rows or columns. The number of selected rows and columns to be inserted must be the same as the dimensions of the insertion parameter, or a scalar that can be repeatedly inserted at the index position.

You can use operators to assign values ​​to matrices or matrix subscripts, such as =

*= A[loc(A<0)]*=-1;

Use the index operator to create a range matrix. The operator is::

The loc function can be used to create a matrix consisting of all positions where A is true. It is a column matrix and the positions are single subscripts.

Used to return the position of the minimum and maximum elements of the matrix. The position is a single subscript representation.

A matrix of positions used to create matrix A where the values ​​at those positions are less than or equal to a given value given in B

Inverse or inv, inverse is used to return the inverse matrix of the square matrix (non-singular matrix parameters), which can also be abbreviated as INV. The product of the matrix and the inverse matrix returns an identity matrix.

Move data from data table into matrix

Move data from matrix to data table

A matrix of values ​​can be extracted from a window report. If the table box is a variable containing a table box reference, get as matrix will create a matrix A consisting of a column vector of values ​​based on that column.

If colbox is a variable that contains a reference to a report table column box, get as matrix will create a matrix A of column vectors of values ​​based on that column.

Suppose you want to perform your own regression calculations without using JMP's built-in tools

Picture object, which can be used to add pictures to frames or display boxes

Used to open picture objects

Adds new frames to animated images. The frame contains nothing until SetPixel is used. An optional duration parameter can be specified, which indicates the frame duration in milliseconds. If the duration parameter is omitted, the duration of the current frame will be used as the duration of the new frame. If there are no other frames, the default duration of 100 milliseconds will be used.

Using the specified dimensions, trim the existing image to create it. If there is no L value, replace the original image.

Repeated filtering affects image quality

Method 1. Despeckle removes pixels that do not blend with surrounding pixels. For example, black pixels wrapped by white pixels are converted into white pixels.

Method 2, edge edge, all objects outside the outline of the object are darkened

Method 3. Enhance Enhancement Reduce the contrast between pixels in noisy images

Method 4, median replaces the color value of each pixel with the median value of surrounding pixels

Method 5, negate replaces the pixel color with its complementary color

Method 6, normalize removes the percentage of top and bottom color values. The color values ​​are then stretched to fill the rest of the image, a process that increases the intensity of the color.

Method 7. Sharpen to make the edges of pixels clearer

Method 8. Contrast Optimize light and dark colors. Larger values ​​will brighten the image

Method 9. Gamma balances the brightness and RGB ratio of the image. Larger values ​​produce lighter images.

Method 10. Reduce noise Denoising Find the minimum and maximum color values ​​and replace them with values ​​consistent with the surrounding pixels. Larger values ​​will create a smoother image.

Method 11. Gaussian blur Gaussian blur blurs pixels with a specified radius. Specifying a larger radius creates a smoother image. You can also specify sigma in JSL. A larger sigma value will make the image smoother.

Method 12. canny edge detection filter to construct edge images

Different methods can be used interchangeably