LIST OF USEFUL COMMANDS FOR YORICK -**---------------- STARTING YORICK AND LITpro --------------------------**- Open an "emacs" window. Inside emacs, type: M-x yorick (for Meta-x, e.g. Alt-x) >>Yeti 6.1.4 ready. Copyright (c) 1996-2005, Eric Thiébaut. >>Copyright (c) 2005. The Regents of the University of California. >>All rights reserved. Yorick 2.1.02 ready. For help type 'help' If not already done, type: > #include "your_path_to/yoga/LITpro/yorick/LITpro.i" Now the routines of LITpro are available. You can start a model fitting! -**------------------ MEMO OF LITpro COMMANDS --------------------------**- -----------SHORT MEMO FOR YORICK USAGE----------- To obtain the DOCUMENTATION of a function: help, ; or about, "function"; (see help, about) If you have any problem, try again: > #include "your_path_to/yoga/LITpro/yorick/LITpro.i" ----------MODEL AND DATA LOADING---------------- > world=load_model(LIT_MODELS+"Obji_model.i"); ------------PLOTS FOR PRACTICE------------------- > plot_uvcoverage, world, newwin=0; > plot_radial, world, "VIS2", newwin=1; > plot_radial, world, "VISAMP"newwin=2; > plot_radial, world, "VISPHI", newwin=3; > plot_radial, world, "T3AMP", newwin=4; > plot_radial, world, "T3PHI", newwin=5; > plot_baselines, world, newwin=0; > plot_uvmap, world, newwin=6; ---------------PARAMETERS------------------------ > set_parameter, world, param_name, value=, vmin=, vmax=, fixed=; > show_parameters, world; ----------------MODEL FITTING------------------- > go_fit, world; > show_fit, world; > chi2_map, world; -**------------------------------------------------------------------------**- -**-------------------- PART ONE: PRACTICE WITH OBJ1 ---------------------**- -**------------------------------------------------------------------------**- AIM: Try to find the best model and best parameters for Obj1.fits data. a) Files of interest: Several scripts of models have been prepared and are available for Obj1.fits: Obj1_uniform_disk.i Obj1_binary_punct.i Obj1_binary_disk.i These scripts, when called for, will automatically load the datafile Obj1.fits b) What to do in yorick: **UNIFORM DISK** - Try the fit of Obj1.fits, with the model of a uniform disk: > world = load_model(LIT_MODELS+"Obj1_uniform_disk.i"); > show_parameters, world; name value prev_val vmin vmax scale fixed units "i": 0.5 0.5 0 1 AUTO 0 "r": 0 0 0 20 AUTO 0 mas "x": 0 0 -40 40 AUTO 0 mas "y": 0 0 -40 40 AUTO 0 mas -This is a description of the parameters of the model: intensity, radius and coordinates. Now, you may use plots routines (see above PLOTS FOR PRACTICE) to visualize the data and the corresponding modelized values. -Then, you can try to fit the model... > go_fit, world; /*run the model fitting*/ > show_fit, world; /*display the result of the fit*/ Number of iterations 41 (Max Number of iterations 200) Final values for fitted parameters and standard deviation: i = 1 +/- 0.0638877 r = 0.69923 +/- 0.00124571 x = -15.363 +/- 1.68589e+07 y = 14.4235 +/- 1.77159e+07 First Chi2 = 214844 - Last Chi2 = 65879.6 Number of degrees of freedom = 241 Confidence Level = 0 --- Correlation matrix --- i r x y i 0.00408 8.31e-14 0.000113 -8.64e-05 r 8.31e-14 1.55e-06 -245 47.4 x 0.000113 -245 2.84e+14 8.89e+13 y -8.64e-05 47.4 8.89e+13 3.14e+14 **BINARY WITH PUNCTUAL COMPONENTS** - Try now, the fit of Obj1.fits, with the model of a punctual binary: > world = load_model(LIT_MODELS+"Obj1_binary_punct.i"); > show_parameters, world; > go_fit, world; /*run the model fitting*/ > show_fit, world; /*display the result of the fit*/ **BINARY WITH UNIFORM DISK COMPONENTS** - Try now, the fit of Obj1.fits, with the model of two uniform disks: > world = load_model(LIT_MODELS+"Obj1_binary_disk.i"); > show_parameters, world; > go_fit, world; /*run the model fitting*/ > show_fit, world; /*display the result of the fit*/ -**------------------------------------------------------------------------**- -**----------------PART TWO: PRACTICE WITH THE PALPEUR--------------------**- -**------------------------------------------------------------------------**- AIM: Try to cope with the issue of local minima, using a "palpeur". When optimized, Chi2 is subject to converge to a local minimum and stop there. The idea is to detect the local minima, and so to reach the global minimum among them. The "palpeur" tries to enhance dimensions and directions of importance in the brightness distribution. For this, we use a binary model of fixed separation (X,Y) on the sky, and we only fit the intensity ratio. It reveals how much the vector of separation has an importance in the model description. For each fixed separation equal to the position on a 2D-sampled map, we can reach a minimal Chi2, fitting only the intensities. The result is a 2D-map of reached Chi2. -**------------------------------------------------------------------------**- -**----------------PART THREE: PRACTICE WITH OBJ2-------------------------**- -**------------------------------------------------------------------------**- AIM: Thanks to expertise of the two previous parts, try to find the best model on a set of real data Obj2.oifits. a) Files of interest: Several scripts of models available: Obj2_uniform_disk.i Obj2_binary_punct.i Obj2_disk_plus_punct.i Obj2_triple_punct.i These scripts, when called for, will automatically load the datafile Obj2.oifits These are real data! b) What to do in yorick: Load a model using one of the following commands: ** UNIFORM DISK ** > world = load_model(LIT_MODELS+"Obj2_uniform_disk.i"); ** PUNCTUAL BINARY ** > world = load_model(LIT_MODELS+"Obj2_binary_punct.i"); ** UNIFORM DISK AND PUNCTUAL OBJECT ** > world = load_model(LIT_MODELS+"Obj2_disk_and_punct.i"); ** TRIPLE SYSTEM OF PUNCTUAL OBJECTS ** > world = load_model(LIT_MODELS+"Obj2_triple_punct.i"); Then you can see/tune the parameters and run the model fitting: > show_parameters, world; > go_fit, world; /*run the model fitting*/ > show_fit, world; /*display the result of the fit*/ > chi2_map, world; /*run the "palpeur"*/