error_element

• type: setup command.
• function: assert a random error defintion for the accelerator.

&error_element
    STRING name = NULL;
    STRING element_type = NULL;
    STRING item = NULL;
    STRING type = "gaussian";
    double amplitude = 0.0;
    double cutoff = 3.0;
    long bind = 1;
    long bind_number = 0;
    longn bind_across_names = 0;
    long post_correction = 0;
    long fractional = 0;
    long additive = 1;
    STRING after = NULL;
    STRING before = NULL;
&end

• name -- The possibly wildcarded name of the elements for which errors are being specified.
• element_type -- An optional, possibly wildcarded string giving the type of elements to which the errors should be applied. E.g., element_type=*MON* would match all beam position monitors. If this item is given, then name may be left blank.
• item -- The parameter of the elements to which the error pertains.
• type -- The type of random distribution to use. May be one of ``uniform'', ``gaussian'', or ``plus_or_minus''. A ``plus_or_minus'' error is equal in magnitude to the amplitude given, with the sign randomly chosen.
• amplitude -- The amplitude of the errors.
• cutoff -- The cutoff for the random distribution.

• bind, bind_number, bind_across_names -- These parameters control ``binding'' of errors among elements, which means assigning the same error contribution to several elements. This occurs if bind is nonzero; if bind is negative, then the sign of the error will alternate between successive elements. bind_number can be used to limit the number of elements bound together. In particular, if bind_number is positive, then a positive value of bind indicates that bind_number successive elements having the same name will have the same error value. Finally, by default, elegant only binds the errors of objects having the same name, even if they are assigned errors by the same error_element command (i.e., through a wildcard name). If bind_across_names is nonzero, then binding is done even for elements with different names.

• post_correction -- A flag indicating whether the errors should be added after orbit, tune, and chromaticity correction.
• fractional -- A flag indicating whether the errors are fractional, in which case the amplitude refers to the amplitude of the fractional error.
• additive -- A flag indicating that the errors should be added to the prior value of the parameter. If zero, then the errors replace the prior value of the parameter.
• after -- The name of an element. If given, the error is applied only to elements that follow the named element in the beamline.
• before -- The name of an element. If given, the error is applied only to elements that precede the named element in the beamline.