python4astronomers.github.com/fitting/low-level.html at master · rillian/python4astronomers.github.com

History

506 lines (457 loc) · 25.9 KB

Raw

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

481

482

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

506

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"

"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<head>

<title>The low-level Sherpa API — Python4Astronomers 2.0 documentation</title>

var DOCUMENTATION_OPTIONS = {

URL_ROOT: '../',

VERSION: '2.0',

COLLAPSE_INDEX: false,

FILE_SUFFIX: '.html',

HAS_SOURCE: true

};

</script>

$(document).ready(function(){

$(".flip0").click(function(){

$(".panel0").slideToggle("normal");

});

$(".flip1").click(function(){

$(".panel1").slideToggle("normal");

});

$(".flip2").click(function(){

$(".panel2").slideToggle("normal");

});

$(".flip3").click(function(){

$(".panel3").slideToggle("normal");

});

$(".flip4").click(function(){

$(".panel4").slideToggle("normal");

});

$(".flip5").click(function(){

$(".panel5").slideToggle("normal");

});

$(".flip6").click(function(){

$(".panel6").slideToggle("normal");

});

$(".flip7").click(function(){

$(".panel7").slideToggle("normal");

});

$(".flip8").click(function(){

$(".panel8").slideToggle("normal");

});

$(".flip9").click(function(){

$(".panel9").slideToggle("normal");

});

</script>

div.panel0,p.flip0

{

font-size: 0.9em;

margin: 0;

padding: 0.1em 0 0.1em 0.5em;

border-bottom: 1px solid #86989B;

}

p.flip0

{

color: white;

font-weight: bold;

background-color: #AFC1C4;

}

div.panel0

{

display:none;

}

div.panel1,p.flip1

{

font-size: 0.9em;

margin: 0;

padding: 0.1em 0 0.1em 0.5em;

border-bottom: 1px solid #86989B;

}

p.flip1

{

color: white;

font-weight: bold;

background-color: #AFC1C4;

}

div.panel1

{

display:none;

}

div.panel2,p.flip2

{

font-size: 0.9em;

margin: 0;

padding: 0.1em 0 0.1em 0.5em;

border-bottom: 1px solid #86989B;

}

p.flip2

{

color: white;

font-weight: bold;

background-color: #AFC1C4;

}

div.panel2

{

display:none;

}

div.panel3,p.flip3

{

font-size: 0.9em;

margin: 0;

padding: 0.1em 0 0.1em 0.5em;

border-bottom: 1px solid #86989B;

}

p.flip3

{

color: white;

font-weight: bold;

background-color: #AFC1C4;

}

div.panel3

{

display:none;

}

div.panel4,p.flip4

{

font-size: 0.9em;

margin: 0;

padding: 0.1em 0 0.1em 0.5em;

border-bottom: 1px solid #86989B;

}

p.flip4

{

color: white;

font-weight: bold;

background-color: #AFC1C4;

}

div.panel4

{

display:none;

}

div.panel5,p.flip5

{

font-size: 0.9em;

margin: 0;

padding: 0.1em 0 0.1em 0.5em;

border-bottom: 1px solid #86989B;

}

p.flip5

{

color: white;

font-weight: bold;

background-color: #AFC1C4;

}

div.panel5

{

display:none;

}

div.panel6,p.flip6

{

font-size: 0.9em;

margin: 0;

padding: 0.1em 0 0.1em 0.5em;

border-bottom: 1px solid #86989B;

}

p.flip6

{

color: white;

font-weight: bold;

background-color: #AFC1C4;

}

div.panel6

{

display:none;

}

div.panel7,p.flip7

{

font-size: 0.9em;

margin: 0;

padding: 0.1em 0 0.1em 0.5em;

border-bottom: 1px solid #86989B;

}

p.flip7

{

color: white;

font-weight: bold;

background-color: #AFC1C4;

}

div.panel7

{

display:none;

}

div.panel8,p.flip8

{

font-size: 0.9em;

margin: 0;

padding: 0.1em 0 0.1em 0.5em;

border-bottom: 1px solid #86989B;

}

p.flip8

{

color: white;

font-weight: bold;

background-color: #AFC1C4;

}

div.panel8

{

display:none;

}

div.panel9,p.flip9

{

font-size: 0.9em;

margin: 0;

padding: 0.1em 0 0.1em 0.5em;

border-bottom: 1px solid #86989B;

}

p.flip9

{

color: white;

font-weight: bold;

background-color: #AFC1C4;

}

div.panel9

{

display:none;

}

</style>

var _gaq = _gaq || [];

_gaq.push(['_setAccount', 'UA-18709417-2']);

_gaq.push(['_trackPageview']);

(function() {

var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true;

ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js';

var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s);

})();

</script>

</head>

<body>

<a class="brand" title="Documentation Home" href="../index.html">python4astronomers</a>

<ul>

</ul>

</div>

<h3>Navigation</h3>

<ul>

</a>

</li>

« previous

</a>

</li>

<li>

<a href="../index.html">Python4Astronomers 2.0 documentation</a>

</li>

<li><a href="fitting.html" accesskey="U">Fitting and Modeling 1-D and 2-D Data</a> »</li>

<li>The low-level Sherpa API</li>

</ul>

</div>

<h1>The low-level Sherpa API<a class="headerlink" href="#the-low-level-sherpa-api" title="Permalink to this headline">¶</a></h1>

Here we repeat our <a class="reference external" href="spectrum.html">fit to the 3C273 spectrum</a>, but

with the low-level API (and without going quite as far in the analysis).

<h2>Explore the Sherpa Object Model<a class="headerlink" href="#explore-the-sherpa-object-model" title="Permalink to this headline">¶</a></h2>

In a new working directory, download a MAST spectrum of <a class="reference download internal" href="../_downloads/3c273.fits"><tt class="xref download docutils literal">3C 273</tt></a>

and start IPython along with the standard imports:

<div class="highlight-python"><div class="highlight"><pre>$ ipython --matplotlib

import numpy as np

import matplotlib.pyplot as plt

</pre></div>

</div>

If you have trouble accessing the spectrum you can download it straight away

using Python:

<div class="highlight-python"><div class="highlight"><pre>from astropy.extern.six.moves.urllib import request

url = 'http://python4astronomers.github.com/_downloads/3c273.fits'

open('3c273.fits', 'wb').write(request.urlopen(url).read())

%ls

</pre></div>

</div>

Import a few Sherpa classes needed to characterize a fit:

<div class="highlight-python"><div class="highlight"><pre>from sherpa.data import Data1D

from sherpa.models import PowLaw1D

from sherpa.stats import Chi2DataVar

from sherpa.optmethods import LevMar

from sherpa.fit import Fit

</pre></div>

</div>

Import the Python FITS reader <a class="reference external" href="http://docs.astropy.org/en/stable/io/fits/">astropy.io.fits</a> and open the spectrum as a table:

<div class="highlight-python"><div class="highlight"><pre>from astropy.io import fits

dat = fits.open('3c273.fits')[1].data

</pre></div>

</div>

Access the <cite>WAVELENGTH</cite> and <cite>FLUX</cite> columns from the pyFITS <tt class="docutils literal">RecArray</tt>. Populate

variables represented as <tt class="docutils literal">wave</tt>, <tt class="docutils literal">flux</tt>, and <tt class="docutils literal">err</tt>. Normalize the flux and assume

uncertainties of 2% of the flux:

<div class="highlight-python"><div class="highlight"><pre>wave = dat.field('WAVELENGTH')

flux = dat.field('FLUX') * 1e14

err = dat.field('FLUX') * 0.02e14

</pre></div>

</div>

Create a Sherpa <tt class="docutils literal">Data1D</tt> data set from the NumPy arrays <tt class="docutils literal">wave</tt>, <tt class="docutils literal">flux</tt>, and

<tt class="docutils literal">err</tt>. The data arrays are accessible from the <tt class="docutils literal">data</tt> object as the attributes

<tt class="docutils literal">x</tt>, <tt class="docutils literal">y</tt>, and <tt class="docutils literal">staterror</tt>:

<div class="highlight-python"><div class="highlight"><pre>data = Data1D('3C 273', wave, flux, err)

print(data)

</pre></div>

</div>

Array access:

<div class="highlight-python"><div class="highlight"><pre>print('x {0}'.format(data.x))

print('y {0}'.format(data.y))

print('err {0}'.format(data.staterror))

</pre></div>

</div>

Define a convenience function <tt class="docutils literal">plot_data</tt> that calls the matplotlib functions

<tt class="docutils literal">plot</tt> and <tt class="docutils literal">errorbar</tt> according to certain criteria. Plot the <tt class="docutils literal">x</tt> and

<tt class="docutils literal">y</tt> arrays using the format specified in the optional argument, <tt class="docutils literal">fmt</tt>.

Clear the plot if the <tt class="docutils literal">clear</tt> argument is <tt class="docutils literal">True</tt>. Add the plot errorbars if

the <tt class="docutils literal">err</tt> array is present. Plot the spectrum by accessing the NumPy arrays

in the Sherpa data set using our new function and its default arguments:

<div class="highlight-python"><div class="highlight"><pre>def plot_data(x, y, err=None, fmt='.', clear=True):

if clear:

plt.clf()

plt.plot(x, y, fmt)

if err is not None:

plt.errorbar(x, y, err, fmt=None, ecolor='b')

plot_data(data.x, data.y, data.staterror)

</pre></div>

</div>

Create a Sherpa power-law model <tt class="docutils literal">pl</tt>. All Sherpa models maintain a tuple of

parameters in <tt class="docutils literal">pars</tt>. Access each of the model’s parameter objects and print

the <tt class="docutils literal">name</tt> and <tt class="docutils literal">val</tt> attributes:

<div class="highlight-python"><div class="highlight"><pre>pl = PowLaw1D('pl')

pl.pars

for par in pl.pars:

print('{0} {1}'.format(par.name, par.val))

print(pl)

</pre></div>

</div>

Set the power-law reference to be 4000 Angstroms and print out the <tt class="docutils literal">PowLaw1D</tt>

object and its parameter information. Each model parameter is accessible as an

attribute its model. For example, the power-law amplitude is referenced with

<tt class="docutils literal">pl.ampl</tt>:

<div class="highlight-python"><div class="highlight"><pre>pl.ref = 4000.

print(pl)

</pre></div>

</div>

Model parameters are themselves class objects:

<div class="highlight-python"><div class="highlight"><pre>print(pl.ampl)

</pre></div>

</div>

Exercise (for the interested reader): Special methods and properties

Wait. Didn’t we just set <tt class="docutils literal">pl.ref</tt> to be an float? How can <tt class="docutils literal">pl.ref</tt> be an

float and a <tt class="docutils literal">Parameter</tt> object?

</div>

Click to Show/Hide Solution <div class="panel0">The answer is that pl.ref is in fact an object, but its model class supports a

special setter method <tt class="docutils literal">__setattr__()</tt> that updates the pl.ref.val attribute

underneath. The <tt class="docutils literal">property</tt> function defines custom getter and setter

functions for a particular class attribute:

<div class="highlight-python"><div class="highlight"><pre>class Parameter(object):

def __init__(self):

# private attribute intended to be reference as 'val'.

self._value = 1.0

def _get_val(self): return self._value

def _set_val(self, value): self._value = value

# setup a 'val' attribute

val = property(_get_val, _set_val)

class Model(object):

def __setattr__(self, name, val):

if isinstance(getattr(self, name, None), Parameter):

getattr(self, name).val = val

else:

object.__setattr__(self, name, val)

def __init__(self):

self.ref = Parameter()

m = Model()

m.ref

m.ref = 4

m.ref

m.ref.val

</pre></div>

</div>

</div>Create a <tt class="docutils literal">Fit</tt> object made up of a Sherpa data set, model, fit statistic, and

optimization method. Fit the spectrum to a power-law with least squares

(Levenberg-Marquardt) using the chi-squared statistic with data variance:

<div class="highlight-python"><div class="highlight"><pre>f = Fit(data, pl, Chi2DataVar(), LevMar())

result = f.fit()

print(result)

# or alternatively

print(result.format())

</pre></div>

</div>

Over-plot the fitted model atop the data points using our convenience function

<tt class="docutils literal">plot_data</tt>. This time calculate the model using the best-fit parameter

values over the <tt class="docutils literal">data.x</tt> and plot using a custom format and indicate

<tt class="docutils literal">clear=False</tt>:

<div class="highlight-python"><div class="highlight"><pre>plot_data(data.x, pl(data.x), fmt="-", clear=False)

</pre></div>

</div>

</div>

<div class="sphinxsidebarwrapper"><h3>Page Contents</h3>

<ul>

<li><a class="reference internal" href="#">The low-level Sherpa API</a><ul>

<li><a class="reference internal" href="#explore-the-sherpa-object-model">Explore the Sherpa Object Model</a></li>

</ul>

</li>

</ul>

<h4>Previous topic</h4>

<a href="image.html"

title="previous chapter">2-D Fitting in Sherpa</a>

<h4>Next topic</h4>

<a href="../astropy/astropy.html"

title="next chapter">Astropy I: core functions</a>

</div>

</div>

<a href="../_sources/fitting/low-level.txt"

rel="nofollow">Page Source</a>

terms of <a rel="license"

href="http://creativecommons.org/licenses/by/3.0/">CC

Attribution 3.0</a>.

Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.2.3.

</footer>

</body>

</html>

pFad - Phone/Frame/Anonymizer/Declutterfier! Saves Data!

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Pfad - The Proxy pFad © 2024 Your Company Name. All rights reserved.

pFad - Phone/Frame/Anonymizer/Declutterfier! Saves Data!

FilesExpand file tree

low-level.html

Latest commit

History

low-level.html

File metadata and controls

Pfad - The Proxy pFad © 2024 Your Company Name. All rights reserved.