def test_init_from_fits():
""" Load FNConstraint from a set of FITS files
"""
# Load
all_fN_cs = FNConstraint.load_defaults()
# Test
assert len(all_fN_cs) == 8
fN_dtype = [fc.fN_dtype for fc in all_fN_cs]
itau = fN_dtype.index('teff')
np.testing.assert_allclose(all_fN_cs[itau].data['TEFF'], 0.19778812)
def test_init_from_fits():
""" Load FNConstraint from a set of FITS files
"""
# Load
all_fN_cs = FNConstraint.load_defaults()
# Test
assert len(all_fN_cs) == 8
fN_dtype = [fc.fN_dtype for fc in all_fN_cs]
itau = fN_dtype.index('teff')
np.testing.assert_allclose(all_fN_cs[itau].data['TEFF'], 0.19778812)
def tst_fn_data(fN_model=None, model_two=None, data_list=None, outfil=None):
""" Make a matplotlib plot like the final figure from P14
See the Notebook for Bokeh plots
Taken from xastropy without actually trying to run
Parameters
----------
fN_model
model_two
data_list
outfil
"""
import matplotlib as mpl
mpl.rcParams['font.family'] = 'STIXGeneral-Regular' # Not for PDF
mpl.rcParams['lines.linewidth'] = 2
from matplotlib import pyplot as plt
#from matplotlib.backends.backend_pdf import PdfPages
# fN data
all_fN_cs = FNConstraint.load_defaults()
# Remove K12
if data_list is None:
fN_cs = [fN_c for fN_c in all_fN_cs if ((fN_c.ref != 'K02') & (fN_c.ref != 'PW09'))]
else:
fN_cs = [fN_c for fN_c in all_fN_cs if fN_c.ref in data_list]
fN_dtype = [fc.fN_dtype for fc in fN_cs]
fig = plt.figure(figsize=(8, 5))
fig.clf()
main = fig.add_axes( [0.1, 0.1, 0.8, 0.8] ) # xypos, xy-size
# f(N) data
main.set_ylabel(r'$\log f(N_{\rm HI})$')
main.set_xlabel(r'$\log N_{\rm HI}$')
main.set_ylim(-25., -9)
for fN_c in fN_cs:
if fN_c.fN_dtype == 'fN':
# Length
ip = range(fN_c.data['NPT'])
val = np.where(fN_c.data['FN'][ip] > -90)[0]
if len(val) > 0:
ipv = np.array(ip)[val]
xval = np.median(fN_c.data['BINS'][:,ipv],0)
xerror = [ fN_c.data['BINS'][1,ipv]-xval, xval-fN_c.data['BINS'][0,ipv] ]
yerror = [ fN_c.data['SIG_FN'][1,ipv], fN_c.data['SIG_FN'][0,ipv] ] # Inverted!
main.errorbar(xval, fN_c.data['FN'][ipv], xerr=xerror, yerr=yerror, fmt='o',
label=fN_c.ref,capthick=2)
main.legend(loc='lower left', numpoints=1)
# Model?
if fN_model is not None:
xplt = 12.01 + 0.01*np.arange(1100)
yplt = fN_model.evaluate(xplt, 2.4)
main.plot(xplt,yplt,'-',color='black')
print(xplt[0],yplt[0])
if model_two is not None:
xplt = 12.01 + 0.01*np.arange(1100)
yplt = model_two.evaluate(xplt, 2.4)
main.plot(xplt,yplt,'-',color='gray')
# Extras
#mpl.rcParams['lines.capthick'] = 2
inset = fig.add_axes( [0.55, 0.6, 0.25, 0.25] ) # xypos, xy-size
inset.set_ylabel('Value') # LHS
inset.xaxis.set_major_locator(plt.FixedLocator(range(5)))
inset.xaxis.set_major_formatter(plt.FixedFormatter(['',r'$\tau_{\rm eff}^{\rm Ly\alpha}$',
r'$\ell(X)_{\rm LLS}$',
r'$\lambda_{\rm mfp}^{912}$', '']))
inset.set_ylim(0., 0.6)
# tau_eff
flg_teff = 1
try:
itau = fN_dtype.index('teff') # Passes back the first one
except:
flg_teff = 0
if flg_teff:
teff=float(fN_cs[itau].data['TEFF'])
D_A = 1. - np.exp(-1. * teff)
SIGDA_LIMIT = 0.1 # Allows for systemtics and b-value uncertainty
sig_teff = np.max([fN_cs[itau].data['SIG_TEFF'], (SIGDA_LIMIT*teff)])
# Plot
inset.errorbar(1, teff, sig_teff, fmt='_', capthick=2)
# Model
if fN_model is not None:
model_teff = tau_eff.lyman_ew(1215.6701*(1+fN_cs[itau].zeval), fN_cs[itau].zeval+0.1,
fN_model, NHI_MIN=fN_cs[itau].data['NHI_MNX'][0],
NHI_MAX=fN_cs[itau].data['NHI_MNX'][1])
inset.plot(1, model_teff, 'ko')
#xdb.set_trace()
## #######
# LLS constraint
flg_LLS = 1
try:
iLLS = fN_dtype.index('LLS') # Passes back the first one
except:
#raise ValueError('fN.data: Missing LLS type')
flg_LLS = 0
if flg_LLS:
inset.errorbar(2, fN_cs[iLLS].data['LX'], yerr=fN_cs[iLLS].data['SIG_LX'],
fmt='_', capthick=2)
# Model
if fN_model != None:
lX = fN_model.calculate_lox(fN_cs[iLLS].zeval,
17.19+np.log10(fN_cs[iLLS].data['TAU_LIM']), 22.)
inset.plot(2, lX, 'ko')
## #######
# MFP constraint
flg_MFP = 1
try:
iMFP = fN_dtype.index('MFP') # Passes back the first one
except:
#raise ValueError('fN.data: Missing MFP type')
flg_MFP = 0
if flg_MFP:
inset2 = inset.twinx()
inset2.errorbar(3, fN_cs[iMFP].data['MFP'], yerr=fN_cs[iMFP].data['SIG_MFP'],
fmt='_', capthick=2)
inset2.set_xlim(0,4)
inset2.set_ylim(0,350)
inset2.set_ylabel('(Mpc)')
# Model
if fN_model is not None:
#fN_model.zmnx = (0.1, 20.) # Reset for MFP calculation
mfp = fN_model.mfp(fN_cs[iMFP].zeval)
inset2.plot(3, mfp, 'ko')
# Show
if outfil != None:
plt.savefig(outfil,bbox_inches='tight')
else:
plt.show()
