Commit 8e6b49b812e99590b90d6b1eaf0b0fd8e67220dd
1 parent
ce9cf8b3
rename to make_unbinned_analysis.py
Showing
1 changed file
with
0 additions
and
427 deletions
examples/unbinned_analysis.py deleted
| 1 | -#! /usr/bin/env python | |
| 2 | -# ========================================================================== | |
| 3 | -# This script illustrates how to build an unbinned analysis pipeline with | |
| 4 | -# the ctatools. If matplotlib is installed, the spectrum will be displayed | |
| 5 | -# on the screen. | |
| 6 | -# | |
| 7 | -# Required 3rd party modules: | |
| 8 | -# None | |
| 9 | -# | |
| 10 | -# Copyright (C) 2011 Jurgen Knodlseder | |
| 11 | -# | |
| 12 | -# This program is free software: you can redistribute it and/or modify | |
| 13 | -# it under the terms of the GNU General Public License as published by | |
| 14 | -# the Free Software Foundation, either version 3 of the License, or | |
| 15 | -# (at your option) any later version. | |
| 16 | -# | |
| 17 | -# This program is distributed in the hope that it will be useful, | |
| 18 | -# but WITHOUT ANY WARRANTY; without even the implied warranty of | |
| 19 | -# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
| 20 | -# GNU General Public License for more details. | |
| 21 | -# | |
| 22 | -# You should have received a copy of the GNU General Public License | |
| 23 | -# along with this program. If not, see <http://www.gnu.org/licenses/>. | |
| 24 | -# | |
| 25 | -# ========================================================================== | |
| 26 | -from ctatools import * | |
| 27 | -from gammalib import * | |
| 28 | -from math import * | |
| 29 | -import os | |
| 30 | -import glob | |
| 31 | -import sys | |
| 32 | - | |
| 33 | - | |
| 34 | -# ============================= # | |
| 35 | -# Analysis pipeline - version 1 # | |
| 36 | -# ============================= # | |
| 37 | -def pipeline_v1(): | |
| 38 | - """ | |
| 39 | - Unbinned analysis pipeline - save intermediate results. | |
| 40 | - | |
| 41 | - This function implements an analysis pipeline that successively calls | |
| 42 | - ctobssim, ctselect and ctlike by saving the intermediate results as FITS | |
| 43 | - files on disk. This replicates the way how the analysis would be done | |
| 44 | - in a ftools-like approach. | |
| 45 | - """ | |
| 46 | - # Set script parameters | |
| 47 | - model_name = "${GAMMALIB}/share/models/crab.xml" | |
| 48 | - events_name = "events.fits" | |
| 49 | - selected_events_name = "selected_events.fits" | |
| 50 | - result_name = "results.xml" | |
| 51 | - caldb = "${GAMMALIB}/share/caldb/cta" | |
| 52 | - irf = "kb_E_50h_v3" | |
| 53 | - ra = 83.63 | |
| 54 | - dec = 22.01 | |
| 55 | - rad_sim = 10.0 | |
| 56 | - tstart = 0.0 | |
| 57 | - tstop = 1800.0 | |
| 58 | - emin = 0.1 | |
| 59 | - emax = 100.0 | |
| 60 | - rad_select = 3.0 | |
| 61 | - | |
| 62 | - # Initialise timing | |
| 63 | - wall_seconds = 0.0 | |
| 64 | - cpu_seconds = 0.0 | |
| 65 | - | |
| 66 | - # Simulate events | |
| 67 | - sim = ctobssim() | |
| 68 | - sim.logFileOpen() | |
| 69 | - sim["infile"].filename(model_name) | |
| 70 | - sim["outfile"].filename(events_name) | |
| 71 | - sim["caldb"].string(caldb) | |
| 72 | - sim["irf"].string(irf) | |
| 73 | - sim["ra"].real(ra) | |
| 74 | - sim["dec"].real(dec) | |
| 75 | - sim["rad"].real(rad_sim) | |
| 76 | - sim["tmin"].real(tstart) | |
| 77 | - sim["tmax"].real(tstop) | |
| 78 | - sim["emin"].real(emin) | |
| 79 | - sim["emax"].real(emax) | |
| 80 | - sim.execute() | |
| 81 | - print "Simulated events ("+str(sim.celapse())+" CPU seconds)" | |
| 82 | - | |
| 83 | - # Update timing | |
| 84 | - wall_seconds += sim.telapse() | |
| 85 | - cpu_seconds += sim.celapse() | |
| 86 | - | |
| 87 | - # Select events | |
| 88 | - select = ctselect() | |
| 89 | - select.logFileOpen() | |
| 90 | - select["infile"].filename(events_name) | |
| 91 | - select["outfile"].filename(selected_events_name) | |
| 92 | - select["ra"].real(ra) | |
| 93 | - select["dec"].real(dec) | |
| 94 | - select["rad"].real(rad_select) | |
| 95 | - select["tmin"].real(tstart) | |
| 96 | - select["tmax"].real(tstop) | |
| 97 | - select["emin"].real(emin) | |
| 98 | - select["emax"].real(emax) | |
| 99 | - select.execute() | |
| 100 | - print "Select event ("+str(select.celapse())+" CPU seconds)" | |
| 101 | - | |
| 102 | - # Update timing | |
| 103 | - wall_seconds += select.telapse() | |
| 104 | - cpu_seconds += select.celapse() | |
| 105 | - | |
| 106 | - # Perform maximum likelihood fitting | |
| 107 | - like = ctlike() | |
| 108 | - like.logFileOpen() | |
| 109 | - like["evfile"].filename(selected_events_name) | |
| 110 | - like["srcmdl"].filename(model_name) | |
| 111 | - like["outmdl"].filename(result_name) | |
| 112 | - like["method"].string("UNBINNED") | |
| 113 | - like["caldb"].string(caldb) | |
| 114 | - like["irf"].string(irf) | |
| 115 | - like.execute() | |
| 116 | - print "Maximum likelihood fitting ("+str(like.celapse())+" CPU seconds)" | |
| 117 | - | |
| 118 | - # Update timing | |
| 119 | - wall_seconds += like.telapse() | |
| 120 | - cpu_seconds += like.celapse() | |
| 121 | - | |
| 122 | - # Show total times | |
| 123 | - print "Total wall time elapsed: "+str(wall_seconds)+" seconds" | |
| 124 | - print "Total CPU time used ...: "+str(cpu_seconds)+" seconds" | |
| 125 | - | |
| 126 | - # Return | |
| 127 | - return | |
| 128 | - | |
| 129 | - | |
| 130 | -# ============================= # | |
| 131 | -# Analysis pipeline - version 2 # | |
| 132 | -# ============================= # | |
| 133 | -def pipeline_v2(show_data): | |
| 134 | - """ | |
| 135 | - Unbinned analysis pipeline - keep intermediate results in memory. | |
| 136 | - | |
| 137 | - This function implements an analysis pipeline that successively calls | |
| 138 | - ctobssim, ctselect and ctlike without saving the intermediate results as | |
| 139 | - FITS files on disk. All data is only hold in memory. | |
| 140 | - """ | |
| 141 | - # Set script parameters | |
| 142 | - model_name = "${GAMMALIB}/share/models/crab.xml" | |
| 143 | - caldb = "${GAMMALIB}/share/caldb/cta" | |
| 144 | - irf = "kb_E_50h_v3" | |
| 145 | - ra = 83.63 | |
| 146 | - dec = 22.01 | |
| 147 | - rad_sim = 10.0 | |
| 148 | - tstart = 0.0 | |
| 149 | - tstop = 1800.0 | |
| 150 | - emin = 0.1 | |
| 151 | - emax = 100.0 | |
| 152 | - rad_select = 3.0 | |
| 153 | - | |
| 154 | - # Initialise timing | |
| 155 | - wall_seconds = 0.0 | |
| 156 | - cpu_seconds = 0.0 | |
| 157 | - | |
| 158 | - # Simulate events | |
| 159 | - sim = ctobssim() | |
| 160 | - sim["infile"].filename(model_name) | |
| 161 | - sim["caldb"].string(caldb) | |
| 162 | - sim["irf"].string(irf) | |
| 163 | - sim["ra"].real(ra) | |
| 164 | - sim["dec"].real(dec) | |
| 165 | - sim["rad"].real(rad_sim) | |
| 166 | - sim["tmin"].real(tstart) | |
| 167 | - sim["tmax"].real(tstop) | |
| 168 | - sim["emin"].real(emin) | |
| 169 | - sim["emax"].real(emax) | |
| 170 | - sim.run() | |
| 171 | - print "Simulated events ("+str(sim.celapse())+" CPU seconds)" | |
| 172 | - | |
| 173 | - # Update timing | |
| 174 | - wall_seconds += sim.telapse() | |
| 175 | - cpu_seconds += sim.celapse() | |
| 176 | - | |
| 177 | - # Select events | |
| 178 | - select = ctselect(sim.obs()) | |
| 179 | - select["ra"].real(ra) | |
| 180 | - select["dec"].real(dec) | |
| 181 | - select["rad"].real(rad_select) | |
| 182 | - select["tmin"].real(tstart) | |
| 183 | - select["tmax"].real(tstop) | |
| 184 | - select["emin"].real(emin) | |
| 185 | - select["emax"].real(emax) | |
| 186 | - select.run() | |
| 187 | - print "Select event ("+str(select.celapse())+" CPU seconds)" | |
| 188 | - | |
| 189 | - # Update timing | |
| 190 | - wall_seconds += select.telapse() | |
| 191 | - cpu_seconds += select.celapse() | |
| 192 | - | |
| 193 | - # Perform maximum likelihood fitting | |
| 194 | - like = ctlike(select.obs()) | |
| 195 | - like.run() | |
| 196 | - print "Maximum likelihood fitting ("+str(like.celapse())+" CPU seconds)" | |
| 197 | - | |
| 198 | - # Update timing | |
| 199 | - wall_seconds += like.telapse() | |
| 200 | - cpu_seconds += like.celapse() | |
| 201 | - | |
| 202 | - # Show total times | |
| 203 | - print "Total wall time elapsed: "+str(wall_seconds)+" seconds" | |
| 204 | - print "Total CPU time used ...: "+str(cpu_seconds)+" seconds" | |
| 205 | - | |
| 206 | - # Show model fitting results | |
| 207 | - #print like.obs().models() | |
| 208 | - | |
| 209 | - # Optionally plot counts | |
| 210 | - if show_data: | |
| 211 | - plot_counts(like.obs()) | |
| 212 | - | |
| 213 | - # Return | |
| 214 | - return | |
| 215 | - | |
| 216 | - | |
| 217 | -# =========== # | |
| 218 | -# Plot counts # | |
| 219 | -# =========== # | |
| 220 | -def plot_counts(observations): | |
| 221 | - """ | |
| 222 | - Plot counts. | |
| 223 | - """ | |
| 224 | - # Only proceed if matplotlib is available | |
| 225 | - try: | |
| 226 | - # Import matplotlib | |
| 227 | - import matplotlib.pyplot as plt | |
| 228 | - | |
| 229 | - # Set legend fontsize | |
| 230 | - params = {'legend.fontsize': 10} | |
| 231 | - plt.rcParams.update(params) | |
| 232 | - | |
| 233 | - # Set plot styles | |
| 234 | - styles = ['b-', 'g-', 'y-', 'n-'] | |
| 235 | - | |
| 236 | - # Dump header | |
| 237 | - print "" | |
| 238 | - print "Make plots (using matplotlib):" | |
| 239 | - print "==============================" | |
| 240 | - | |
| 241 | - # Create figure 1 | |
| 242 | - plt.figure(1,figsize=(12,6)) | |
| 243 | - plt.subplots_adjust(hspace=.7) | |
| 244 | - | |
| 245 | - # Create subplot 1 | |
| 246 | - plt.subplot(121) | |
| 247 | - plt.title("Spectrum (summed over all pixels)") | |
| 248 | - | |
| 249 | - # Loop over observations | |
| 250 | - for obs in observations: | |
| 251 | - | |
| 252 | - # Get event list | |
| 253 | - list = cast_GCTAEventList(obs.events()) | |
| 254 | - | |
| 255 | - # Create energy axis | |
| 256 | - ebounds = GEbounds() | |
| 257 | - emin = GEnergy() | |
| 258 | - emax = GEnergy() | |
| 259 | - emin.TeV(0.1) | |
| 260 | - emax.TeV(100.0) | |
| 261 | - ebounds.setlog(emin, emax, 10) | |
| 262 | - energy = [ebounds.elogmean(i).TeV() for i in range(ebounds.size())] | |
| 263 | - | |
| 264 | - # Create spectrum | |
| 265 | - print "Extract data:" | |
| 266 | - counts = [0.0 for i in range(ebounds.size())] | |
| 267 | - print list.size() | |
| 268 | - for atom in list: | |
| 269 | - index = ebounds.index(atom.energy()) | |
| 270 | - counts[index] = counts[index] + 1.0 | |
| 271 | - | |
| 272 | - # Create error bars | |
| 273 | - error = [sqrt(c) for c in counts] | |
| 274 | - | |
| 275 | - # Plot spectrum | |
| 276 | - plt.loglog(energy, counts, 'ro', label='data') | |
| 277 | - #plt.errorbar(energy, counts, error, fmt=None, ecolor='r') | |
| 278 | - | |
| 279 | - # Extract models | |
| 280 | - print "Extract models:" | |
| 281 | - sum_model = [0.0 for i in range(ebounds.size())] | |
| 282 | - for k, m in enumerate(observations.models()): | |
| 283 | - print "- "+m.name() | |
| 284 | - model = [0.0 for i in range(ebounds.size())] | |
| 285 | - for atom in list: | |
| 286 | - index = ebounds.index(atom.energy()) | |
| 287 | - prob = m.eval(atom, obs) | |
| 288 | - model[index] = model[index] + prob * atom.size() | |
| 289 | - for i in range(ebounds.size()): | |
| 290 | - sum_model[i] = sum_model[i] + model[i] | |
| 291 | - #plt.loglog(energy, model, styles[k], label=m.name()) | |
| 292 | - #plt.loglog(energy, sum_model, 'r-', label='total') | |
| 293 | - | |
| 294 | - # Put labels | |
| 295 | - plt.xlabel("Energy (TeV)") | |
| 296 | - plt.ylabel("Counts") | |
| 297 | - plt.legend(loc="lower left") | |
| 298 | - | |
| 299 | - # Create subplot 2 | |
| 300 | - plt.subplot(122) | |
| 301 | - plt.title("Offset (summed over all energies)") | |
| 302 | - | |
| 303 | - # Set Crab direction | |
| 304 | - crab = GSkyDir() | |
| 305 | - crab.radec_deg(83.63, 22.01) | |
| 306 | - | |
| 307 | - # Loop over observations | |
| 308 | - for obs in observations: | |
| 309 | - | |
| 310 | - # Get event cube | |
| 311 | - list = cast_GCTAEventList(obs.events()) | |
| 312 | - | |
| 313 | - # Create offset histogram | |
| 314 | - print "Extract data:" | |
| 315 | - nx = 30 | |
| 316 | - doffset2 = 0.01 | |
| 317 | - offset2 = [(i+0.5)*doffset2 for i in range(nx)] | |
| 318 | - counts = [0.0 for i in range(nx)] | |
| 319 | - for atom in list: | |
| 320 | - off = atom.dir().dist_deg(crab) | |
| 321 | - off2 = off*off | |
| 322 | - index = int(off2/doffset2) | |
| 323 | - if index < nx: | |
| 324 | - counts[index] = counts[index] + 1.0 | |
| 325 | - | |
| 326 | - # Create error bars | |
| 327 | - error = [sqrt(c) for c in counts] | |
| 328 | - | |
| 329 | - # Plot distribution | |
| 330 | - plt.semilogy(offset2, counts, 'ro', label='data') | |
| 331 | - | |
| 332 | - # Extract models | |
| 333 | - print "Extract models:" | |
| 334 | - sum_model = [0.0 for i in range(nx)] | |
| 335 | - for k, m in enumerate(observations.models()): | |
| 336 | - print "- "+m.name() | |
| 337 | - model = [0.0 for i in range(nx)] | |
| 338 | - for atom in list: | |
| 339 | - off = atom.dir().dist_deg(crab) | |
| 340 | - off2 = off*off | |
| 341 | - index = int(off2/doffset2) | |
| 342 | - if index < nx: | |
| 343 | - prob = m.eval(atom, obs) | |
| 344 | - model[index] = model[index] + prob * atom.size() | |
| 345 | - for i in range(nx): | |
| 346 | - sum_model[i] = sum_model[i] + model[i] | |
| 347 | - #plt.plot(offset2, model, styles[k], label=m.name()) | |
| 348 | - #plt.plot(offset2, sum_model, 'r-', label='total') | |
| 349 | - #plt.ylim(ymin=0.1) | |
| 350 | - | |
| 351 | - # Put labels | |
| 352 | - plt.xlabel("Offset (deg^2)") | |
| 353 | - plt.ylabel("Counts") | |
| 354 | - plt.legend(loc="upper right") | |
| 355 | - | |
| 356 | - # Show counts spectra | |
| 357 | - plt.show() | |
| 358 | - | |
| 359 | - except ImportError: | |
| 360 | - print "Matplotlib is not (correctly) installed on your system. No counts spectra are shown." | |
| 361 | - | |
| 362 | - # Return | |
| 363 | - return | |
| 364 | - | |
| 365 | - | |
| 366 | -#==========================# | |
| 367 | -# Main routine entry point # | |
| 368 | -#==========================# | |
| 369 | -if __name__ == '__main__': | |
| 370 | - """ | |
| 371 | - CTA unbinned analysis script. | |
| 372 | - | |
| 373 | - This script implements the analysis steps to perform an unbinned analysis. | |
| 374 | - The steps include: | |
| 375 | - - Event simulation | |
| 376 | - - Event selection | |
| 377 | - - Binned maximum likelihood fitting | |
| 378 | - | |
| 379 | - Two variations of the script are implemented. One that calls the classes | |
| 380 | - like executables, storing the intermediate results on disk, and a second | |
| 381 | - that uses the results of the precedent class directly without storing | |
| 382 | - any data on disk. | |
| 383 | - """ | |
| 384 | - # Initialise flags | |
| 385 | - need_help = False | |
| 386 | - show_data = True | |
| 387 | - | |
| 388 | - # Test for command line arguments | |
| 389 | - print sys.argv[0] | |
| 390 | - if (len(sys.argv) > 1): | |
| 391 | - if sys.argv[1] == "-h": | |
| 392 | - need_help = True | |
| 393 | - elif sys.argv[1] == "-noshow": | |
| 394 | - show_data = False | |
| 395 | - else: | |
| 396 | - need_help = True | |
| 397 | - | |
| 398 | - # Print help if needed and exit | |
| 399 | - if need_help: | |
| 400 | - print "Usage: example_binned.py [OPTIONS]" | |
| 401 | - print " -h Display this usage message" | |
| 402 | - print " -noshow Do not show data using matplotlib" | |
| 403 | - sys.exit() | |
| 404 | - | |
| 405 | - # Dump header | |
| 406 | - print "*********************************" | |
| 407 | - print "* CTA unbinned analysis scripts *" | |
| 408 | - print "*********************************" | |
| 409 | - | |
| 410 | - # Remove any existing result files | |
| 411 | - list = [glob.glob("*.fits"), glob.glob("*.log"), glob.glob("*.xml")] | |
| 412 | - for files in list: | |
| 413 | - for file in files: | |
| 414 | - os.remove(file) | |
| 415 | - | |
| 416 | - # Save intermediate results on disk | |
| 417 | - print "" | |
| 418 | - print "Executable analysis pipeline:" | |
| 419 | - print "=============================" | |
| 420 | - pipeline_v1() | |
| 421 | - | |
| 422 | - # Do analysis in memory | |
| 423 | - print "" | |
| 424 | - print "In memory analysis pipeline:" | |
| 425 | - print "============================" | |
| 426 | - pipeline_v2(show_data) | |
| 427 | - print "" |