Institut für Astronomie und AstrophysikAbteilung AstronomieSand 1, D-72076 Tübingen, Germany |
epfold
performs epoche folding on data set(s) withing certain period ranges. Tries to look for periods.
cafe
epfold[,range][,subgroup][,group][,/add]
range - (optional) Defines range of data points to fold. This can be either: - Data point range, denoted by the data point number: <n1>[-], while , are positive numbers. Data point numbers start from zero. Open intervals can be represented with a "*" instead of the number. If only one number is given, a single data point will be referenced. If <val2> is less than all but the interval <val2>.. are referenced. - X value range, denoted by float point numbers: <val1>- , while , represents X values defining the interval to reference. Open intervals can be represented with a * instead of the value. If <val2> is less than <val1> all but the interval <val2>.. is referenced. - Vector of data point indices. The vector is denoted with [val,val,...,val] (indgen may be used also). - Boolean expressions with X/Y values. All values for which this expression is true are referenced. The expression may contain algebraic formulas combined with comparison operators "lt", "le", "eq", "gt", "ge" as used in IDL comparisons. For each data point a keyword is defined which has meaning and can be used for the boolean expression: x - The x-column value. y - The y-column value. error - The error column value. selected - True when the data point is selected. def - True when the data point is defined. x1 - Multi column first column value. x2 - Multi column second column value. x3 - Multi column third column value. To avoid interference with functions which also can be used a lower/upper case distinction is made. IDL funktions should be upper case. Default is the entire data set to fold. subgroup - (optional) The data set (subgroup) for which the model should be applied. This can be either the subgroup number or the file name representing the data set. It is possible to set more than one subgroup, either with numbers in brackets ([]) separated with ","; or denoting filenames with wildcards ("*"). Default are all subgroups in currently used group. group - (optional) The data group for which folding should applied. Default is the current group. Must be in range [0..29].
pmin - Lowest period for range. Default: 1 pmax - Highest period for range. Default: 100 nbins - Number of bins to use for folding. Default: 20. sampling - how many periods per peak to use Default:10. linear - if set, use a linear trial period array with step equal to the value of linear tolerance - parameter defining the lower limit for the gap length; the reference is the time difference between the first and second entry in the time array; tolerance defines the maximum allowed relative deviation from this reference bin length; default: 1e-8; this parameter is passed to timegap (see timgap.pro for further explanation). lstat - Use L-Statistic instead of Chi^2 for chierg[1,*]; Davies, S.R., 1990, MNRAS 244, 93 events - Consider input to contain events only. In this case only the time (x) axis is used and folded as single events. To work properly the data sets must not contain any gaps which are not acknowledged by to be outside of the good time interval (s.b.). gti - Good time interval (for events only). Must be 2-D array with of the form: gti[2,*]. It defines the start (gti[0,*]) and stop (gti[1,*]) time of event times (x) to be used for the epfold process. If not given it will be computed under following assumptions: 1.) start and stop of each data set define the outer gti boundary. 2.) ignored data points define inner gti boundaries. pdot - Take period derivative into account. If a single value is given this one will be computed. If pdot is of the form pdot = [a,b] then the period derivative will be scanned from in the range a..b. The number of steps are defined with pdotnum (s.b.). pdotnum - Number of steps to perform if pdot defines a scanning interval a..b. pddot - Second derivative of period to be taken into account. time0 - Epoch for folding, needed when usind pdot/pddot. verbose - Show progress information.
All data points given as input are summed up within a histogram according their x value modulo a period. The period must be set before. Dataset name/File name of the product are "EPFOLD", if more than one pdot is scanned (pdotnum>1) the dataset name is "EPFOLD_PDOT".
Creates/Overrides first subgroup in the result group.
The input lightcurve has to be given in count rates (not in photon numbers). To prevent convergence problems of the fit (only if keyword /fitchi is set), the maximum of the chi^2 distribution has to be in between pmin and pmax, and the period interval has to be resonable large.
The command calls the procedure epfold.pro. This subroutine performs a period search using epoch folding. For each trial period, a pulse profile is generated using pfold. This profile is then tested for constancy using a chi^2 test. The maximum chi^2, i.e., the maximum deviation, is the most possible period. This is done for all periods between pmin and pmax using a grid-search operating on the minimum possible period that can still be detected (given by p^2/t). Caveat: The significance of the found periods SHOULD NOT be tested using the chi^2 value obtained from this distributed (see Davies, 1990, Schwarzenberg-Czerny, 1989, Larsson, 1996) Read (and understand) the references before using this routine!
Davies, S.R., 1990, MNRAS 244, 93 Larsson, S., 1996, A&AS 117, 197 Leahy, D.A., Darbro, W., Elsner, R.F., et al., 1983, ApJ 266, 160-170 Schwarzenberg-Czerny, A., 1989, MNRAS 241, 153
> set, pmin=20 > set, pmax=300 > epfold -> Search periods within a range from periods=20..300 for -> data in current group.
$Id: cafe_epfold.pro,v 1.16 2005/02/15 10:40:40 goehler Exp $
[Home Page] [Software, Documentation] [IDL Documentation] [Quick Reference] [Feedback]