I'm not sure that putting summary and raw data in the same table is necessarily a good idea. But, if you need to for some reason (and if I have understood correctly), then you can use 'Tables > Join' with 'By Row Number'. In JSL:
Names Default To Here( 1 );
// First table
dt1 = Open( "$SAMPLE_DATA/Boston Housing.jmp" );
// Summary table
dt2 = dt1 << Summary(
Group( :chas, :tax ),
Mean( :mvalue ),
Subgroup( :radial ),
Freq( "None" ),
Weight( "None" )
);
// Joined table
dt3 = dt1 << Join( With( dt2 ), By Row Number, Output Table( "Joined Data" ) );