cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Check out the JMP® Marketplace featured Capability Explorer add-in
Choose Language Hide Translation Bar
BenGengenbach
Level III

Can I modify "Remembered settings" in a table and re-import them into the profiler?

Hi JMP community,

 

I would like to modify and run different settings of the prediction profiler in a more efficient way.

Modifying each single entry by hand is tedious.

 

Can I modify the settings in an exported table and re-impiort them into the profiler to predict the new settings?

I am alos open to other suggestions.

 

cheers 

ben

 

1 ACCEPTED SOLUTION

Accepted Solutions

Re: Can I modify "Remembered settings" in a table and re-import them into the profiler?

Hi @BenGengenbach ,

It seems that you need to use script to do what you want.
Here is a sample script. I hope it works for you.

1. Run fit model
2. Open factor setting table you want to apply
3. Open script file and run script

 

rpt = Current Report();
obj = rpt << XPath( "//OutlineBox[@helpKey='Profiler']" );
profiler = obj[1] << get scriptable object;
current_setting_values = profiler << Get Factor Settings;

//Select data table
dtNames = {};
For( t = 1, t <= N Table(), t++,
	Insert Into( dtNames, Data Table( t ) << getName )
);
nw2 = New Window( "Data table selector",
	<<modal(),
	<<Return Result,
	H List Box(
		Panel Box( "Pick a data table",
			dlg_dtlb = List Box( dtNames, width( 280 ), nlines( 5 ), MaxItems( 1 ) ), 
		)
	)
);

//Set the factor values and remember 
setting_tb = Data Table( nw2["dlg_dtlb"][1] );
new_settings_list = {};
fac_name_list = {};
new_settings_script = "";

For( s = 1, s <= N Items( current_setting_values ), s++,
	isstring1 = Is Missing( Min( setting_tb[0, 1 + s] ) );
	isstring2 = Is Missing( Min( Num( setting_tb[0, 1 + s][1] ) ) );
	If( isstring1,
		Insert Into( new_settings_list, Eval List( {setting_tb[0, 1 + s]} ) ),
		Insert Into( new_settings_list, setting_tb[0, 1 + s] );
	
	);
	If( isstring2,
		new_settings_script = new_settings_script || Regex(
			Char( current_setting_values[s] ), 
			"(.*?) = (.*?)",
			"\1"
		) || "( \!"_val_" || Char( s ) || "\!", Lock(0), Show(1)),"
	,
		new_settings_script = new_settings_script || Regex(
			Char( current_setting_values[s] ), 
			"(.*?) = (.*?)",
			"\1"
		) || "( _val_" || Char( s ) || ", Lock(0), Show(1)),"
	)
	;
);

new_settings_script = "profiler << Term Value(" || new_settings_script || ")";
For( i = 1, i <= N Row( setting_tb ), i++,
	script = new_settings_script;
	For( f = 1, f <= N Items( current_setting_values ), f++,
		Substitute Into( script, "_val_" || Char( f ), Char( new_settings_list[f][i] ) )
	);
	Try(
		Eval( Parse( script ) );
		profiler << Remember Settings;
	);
);

//Set Setting name to the report
jmp_version = JMP Version();
jmp_version_num = Num( Left( jmp_version, Contains( jmp_version, "." ) - 1 ) );

if(jmp_version_num >= 17,
s_stringcolbox = obj[1][Outline Box( 2 )][Table Box( 1 )][String Col Edit Box( 1 )];,
s_stringcolbox = obj[1][Outline Box( 2 )][Table Box( 1 )][String Col Box( 1 )];
);
setting_name = s_stringcolbox << get();
new_settings_name = setting_tb[0, 1];
m = N Items( setting_name ) - N Items( new_settings_name ) + 1;
n = N Items( new_settings_name );
j = 1;
For( i = m, i <= N Items( setting_name ), i++,
	setting_name[i] = new_settings_name[j];
	j = j + 1;
);
s_stringcolbox << set values( setting_name ); 

Unfortunately, I could not find a way to delete the current settings with a script, so you have to do it manually.

 

View solution in original post

4 REPLIES 4

Re: Can I modify "Remembered settings" in a table and re-import them into the profiler?

Hi @BenGengenbach ,

It seems that you need to use script to do what you want.
Here is a sample script. I hope it works for you.

1. Run fit model
2. Open factor setting table you want to apply
3. Open script file and run script

 

rpt = Current Report();
obj = rpt << XPath( "//OutlineBox[@helpKey='Profiler']" );
profiler = obj[1] << get scriptable object;
current_setting_values = profiler << Get Factor Settings;

//Select data table
dtNames = {};
For( t = 1, t <= N Table(), t++,
	Insert Into( dtNames, Data Table( t ) << getName )
);
nw2 = New Window( "Data table selector",
	<<modal(),
	<<Return Result,
	H List Box(
		Panel Box( "Pick a data table",
			dlg_dtlb = List Box( dtNames, width( 280 ), nlines( 5 ), MaxItems( 1 ) ), 
		)
	)
);

//Set the factor values and remember 
setting_tb = Data Table( nw2["dlg_dtlb"][1] );
new_settings_list = {};
fac_name_list = {};
new_settings_script = "";

For( s = 1, s <= N Items( current_setting_values ), s++,
	isstring1 = Is Missing( Min( setting_tb[0, 1 + s] ) );
	isstring2 = Is Missing( Min( Num( setting_tb[0, 1 + s][1] ) ) );
	If( isstring1,
		Insert Into( new_settings_list, Eval List( {setting_tb[0, 1 + s]} ) ),
		Insert Into( new_settings_list, setting_tb[0, 1 + s] );
	
	);
	If( isstring2,
		new_settings_script = new_settings_script || Regex(
			Char( current_setting_values[s] ), 
			"(.*?) = (.*?)",
			"\1"
		) || "( \!"_val_" || Char( s ) || "\!", Lock(0), Show(1)),"
	,
		new_settings_script = new_settings_script || Regex(
			Char( current_setting_values[s] ), 
			"(.*?) = (.*?)",
			"\1"
		) || "( _val_" || Char( s ) || ", Lock(0), Show(1)),"
	)
	;
);

new_settings_script = "profiler << Term Value(" || new_settings_script || ")";
For( i = 1, i <= N Row( setting_tb ), i++,
	script = new_settings_script;
	For( f = 1, f <= N Items( current_setting_values ), f++,
		Substitute Into( script, "_val_" || Char( f ), Char( new_settings_list[f][i] ) )
	);
	Try(
		Eval( Parse( script ) );
		profiler << Remember Settings;
	);
);

//Set Setting name to the report
jmp_version = JMP Version();
jmp_version_num = Num( Left( jmp_version, Contains( jmp_version, "." ) - 1 ) );

if(jmp_version_num >= 17,
s_stringcolbox = obj[1][Outline Box( 2 )][Table Box( 1 )][String Col Edit Box( 1 )];,
s_stringcolbox = obj[1][Outline Box( 2 )][Table Box( 1 )][String Col Box( 1 )];
);
setting_name = s_stringcolbox << get();
new_settings_name = setting_tb[0, 1];
m = N Items( setting_name ) - N Items( new_settings_name ) + 1;
n = N Items( new_settings_name );
j = 1;
For( i = m, i <= N Items( setting_name ), i++,
	setting_name[i] = new_settings_name[j];
	j = j + 1;
);
s_stringcolbox << set values( setting_name ); 

Unfortunately, I could not find a way to delete the current settings with a script, so you have to do it manually.

 

BenGengenbach
Level III

Re: Can I modify "Remembered settings" in a table and re-import them into the profiler?

Hi @yuichi_katsumur ,

 

wow, thank you for your effort, that is awesome and works nicely!

 

One minor thing: The settings are correctly loaded into the model however I get this error.

 

Cannot subscript Display Box in access or evaluation of 'obj[1][Outline Box(2)][Table Box(1)][ /*###*/String Col Edit Box(1)]' , obj[1][Outline Box( 2 )][Table Box( 1 )][/*###*/String Col Edit Box( 1 )]

 

BenGengenbach_0-1704187187345.png

I am not familiar with JSL so maybe you can help me out here

 

cheers ben

 

Re: Can I modify "Remembered settings" in a table and re-import them into the profiler?

Hello @BenGengenbach ,

Thank you for your reply. I have modified my jsl so that the error does not occur. Please see my jsl in the previous reply.

BenGengenbach
Level III

Re: Can I modify "Remembered settings" in a table and re-import them into the profiler?

Works like a charm!