def augment(xi):
ks_pip_p = np.sqrt(np.sum(xi[:, 0:3]**2, axis=-1)).reshape(-1, 1)
ks_pim_p = np.sqrt(np.sum(xi[:, 3:6]**2, axis=-1)).reshape(-1, 1)
ks_p = np.sqrt(np.sum(xi[:, 7:10]**2, axis=-1)).reshape(-1, 1)
ks_m = np.sqrt(xi[:, 6]**2 - np.sum(xi[:, 7:10]** 2, axis=-1)).reshape(-1, 1)
phi_pip_p = np.sqrt(np.sum(xi[:, 10:13]**2, axis=-1)).reshape(-1, 1)
phi_pim_p = np.sqrt(np.sum(xi[:, 13:16]**2, axis=-1)).reshape(-1, 1)
phi_p = np.sqrt(np.sum(xi[:, 17:20]**2, axis=-1)).reshape(-1, 1)
phi_m = np.sqrt(xi[:, 16]**2 - np.sum(xi[:, 17:20]** 2, axis=-1)).reshape(-1, 1)
d0_p = np.sqrt(np.sum(xi[:, 21:24]**2, axis=-1)).reshape(-1, 1)
d0_m = np.sqrt(xi[:, 20]**2 - np.sum(xi[:, 21:24]** 2, axis=-1)).reshape(-1, 1)
ks_dpx = (xi[:, 0] + xi[:, 3] - xi[:, 7]).reshape(-1, 1)
ks_dpy = (xi[:, 1] + xi[:, 4] - xi[:, 8]).reshape(-1, 1)
ks_dpz = (xi[:, 2] + xi[:, 5] - xi[:, 9]).reshape(-1, 1)
ks_dpe = (energy(MASS_DICT['pi+'], xi[:, 0:3]) + energy(MASS_DICT['pi+'], xi[:, 3:6]) - xi[:, 6]).reshape(-1, 1)
phi_dpx = (xi[:, 10] + xi[:, 13] - xi[:, 17]).reshape(-1, 1)
phi_dpy = (xi[:, 11] + xi[:, 14] - xi[:, 18]).reshape(-1, 1)
phi_dpz = (xi[:, 12] + xi[:, 15] - xi[:, 19]).reshape(-1, 1)
phi_dpe = (energy(MASS_DICT['pi+'], xi[:, 10:13]) + energy(MASS_DICT['pi+'], xi[:, 13:16]) - xi[:, 16]).reshape(-1, 1)
d0_dpx = (xi[:, 7] + xi[:, 17] - xi[:, 21]).reshape(-1, 1)
d0_dpy = (xi[:, 8] + xi[:, 18] - xi[:, 22]).reshape(-1, 1)
d0_dpz = (xi[:, 9] + xi[:, 19] - xi[:, 23]).reshape(-1, 1)
d0_dpe = (xi[:, 6] + xi[:, 16] - xi[:, 20]).reshape(-1, 1)
return np.hstack((xi, ks_pip_p, ks_pim_p, ks_p, ks_m, phi_pip_p, phi_pim_p, phi_p, phi_m, d0_p, d0_m,
ks_dpx, ks_dpy, ks_dpz, ks_dpe, phi_dpx, phi_dpy, phi_dpz, phi_dpe, d0_dpx, d0_dpy, d0_dpz, d0_dpe))
def augment(xi):
ks_m = np.sqrt(xi[:, 6]**2 - np.sum(xi[:, 7:10]
** 2, axis=-1)).reshape(-1, 1)
ks_p = np.sqrt(np.sum(xi[:, 7:10]**2, axis=-1)).reshape(-1, 1)
pip_p = np.sqrt(np.sum(xi[:, :3]**2, axis=-1)).reshape(-1, 1)
pim_p = np.sqrt(np.sum(xi[:, 3:6]**2, axis=-1)).reshape(-1, 1)
dpx = (xi[:, 0] + xi[:, 3] - xi[:, 7]).reshape(-1, 1)
dpy = (xi[:, 1] + xi[:, 4] - xi[:, 8]).reshape(-1, 1)
dpz = (xi[:, 2] + xi[:, 5] - xi[:, 9]).reshape(-1, 1)
dpe = (energy(MASS_DICT['pi+'], xi[:, 0:3]) + energy(MASS_DICT['pi+'], xi[:, 3:6]) -
xi[:, 6]).reshape(-1, 1)
return np.hstack((xi[:, :10], ks_m, ks_p, pip_p, pim_p, dpx, dpy, dpz, dpe))
def plot_pool(xi, cov, pimgen, pipgen):
ks3 = pipgen + pimgen
xi0 = np.hstack((pipgen, pimgen, energy(MASS_DICT['K0_S'],
ks3).reshape(-1, 1), ks3))
label = [
'pip_px', 'pip_py', 'pip_pz', 'pim_px', 'pim_py', 'pim_pz', 'ks_E',
'ks_px', 'ks_py', 'ks_pz'
]
for i in range(xi.shape[1]):
m = (xi[:, i] - xi0[:, i]) / cov[:, i, i]**0.5
for _ in range(5):
mean, std = np.mean(m), np.std(m)
m = m[np.abs(m - mean) < 5. * std]
print(m.shape, np.mean(m), np.std(m))
if np.std(m) < 0.000001:
return
x, bins, e = make_hist(m)
plt.figure(figsize=(6, 5))
plt.errorbar(x, bins, e, linestyle='none', marker='.', markersize=4)
plt.grid()
plt.xlabel(label[i], fontsize=16)
plt.tight_layout()
plt.savefig('fig/pool_{}.png'.format(label[i]))
plt.show()
def plot_hist(xi, xi0, pimgen, pipgen):
ks3 = pipgen + pimgen
xi_gen = np.hstack((pipgen, pimgen, energy(MASS_DICT['K0_S'],
ks3).reshape(-1, 1), ks3))
def augment(xi):
ks_m = np.sqrt(xi[:, 6]**2 - np.sum(xi[:, 7:10]**2, axis=-1)).reshape(
-1, 1)
ks_p = np.sqrt(np.sum(xi[:, 7:10]**2, axis=-1)).reshape(-1, 1)
pip_p = np.sqrt(np.sum(xi[:, :3]**2, axis=-1)).reshape(-1, 1)
pim_p = np.sqrt(np.sum(xi[:, 3:6]**2, axis=-1)).reshape(-1, 1)
dpx = (xi[:, 0] + xi[:, 3] - xi[:, 7]).reshape(-1, 1)
dpy = (xi[:, 1] + xi[:, 4] - xi[:, 8]).reshape(-1, 1)
dpz = (xi[:, 2] + xi[:, 5] - xi[:, 9]).reshape(-1, 1)
return np.hstack((xi, ks_m, ks_p, pip_p, pim_p, dpx, dpy, dpz))
xi = augment(xi)
xi0 = augment(xi0)
xi_gen = augment(xi_gen)
label = [
'pip_px', 'pip_py', 'pip_pz', 'pim_px', 'pim_py', 'pim_pz', 'ks_E',
'ks_px', 'ks_py', 'ks_pz', 'ks_m', 'ks_p', 'pip_p', 'pim_p', 'dpx',
'dpy', 'dpz'
]
for i in range(xi.shape[1]):
fitted = xi[:, i] - xi_gen[:, i]
unfitted = xi0[:, i] - xi_gen[:, i]
for _ in range(5):
fit_mean, fit_std = np.mean(fitted), np.std(fitted)
fitted = fitted[np.abs(fitted - fit_mean) < 5. * fit_std]
unfit_mean, unfit_std = np.mean(unfitted), np.std(unfitted)
unfitted = unfitted[np.abs(unfitted - unfit_mean) < 5. * unfit_std]
plt.figure(figsize=(8, 6))
if len(fitted) != 0:
plt.errorbar(*make_hist(fitted, density=True),
linestyle='none',
marker='.',
markersize=4,
label='fit')
if len(unfitted) != 0:
plt.errorbar(*make_hist(unfitted, density=True),
linestyle='none',
marker='.',
markersize=4,
label='unfit')
plt.plot([], [], ' ', label="fit std {:0.3f}".format(fit_std))
plt.plot([], [], ' ', label="unfit std {:0.3f}".format(unfit_std))
plt.plot([], [], ' ', label="fit mean {:0.3f}".format(fit_mean))
plt.plot([], [], ' ', label="unfit mean {:0.3f}".format(unfit_mean))
plt.legend(loc='upper right')
plt.grid()
plt.xlabel(label[i], fontsize=16)
plt.tight_layout()
plt.savefig('fig/fit_{}.png'.format(label[i]))
plt.show()
def plot_pull_d0(xi, cov, p3_ks_pip_gen, p3_ks_pim_gen, p3_phi_pip_gen, p3_phi_pim_gen, savedir):
ks3 = p3_ks_pip_gen + p3_ks_pim_gen
phi3 = p3_phi_pip_gen + p3_phi_pim_gen
d03 = ks3 + phi3
eks_gen = energy(MASS_DICT['K0_S'], ks3).reshape(-1, 1)
ephi_gen = energy(MASS_DICT['phi'], phi3).reshape(-1, 1)
ed0_gen = energy(MASS_DICT['D0'], d03).reshape(-1, 1)
xi_gen = np.hstack((p3_ks_pip_gen, p3_ks_pim_gen, eks_gen, ks3,
p3_phi_pip_gen, p3_phi_pim_gen, ephi_gen, phi3,
ed0_gen, d03))
if (savedir.split('/')[0] == 'reffit'):
xi = np.hstack((xi[:, 0:10], xi[:,15:25], xi[:,30:34]))
cov[:, 10:20, 10:20] = cov[:, 15:25, 15:25]
cov[:, 20:24, 20:24] = cov[:, 30:34, 30:34]
filenames = ['ks_pip_px', 'ks_pip_py', 'ks_pip_pz', 'ks_pim_px', 'ks_pim_py', 'ks_pim_pz', 'ks_E', 'ks_px', 'ks_py', 'ks_pz',
'phi_pip_px', 'phi_pip_py', 'phi_pip_pz', 'phi_pim_px', 'phi_pim_py', 'phi_pim_pz', 'phi_E', 'phi_px', 'phi_py', 'phi_pz',
'd0_E', 'd0_px', 'd0_py', 'd0_pz',
]
labels = { 'ks_pip_px':r'$p_x(\pi^+_{K_S^0})$ (MeV)',
'ks_pip_py':r'$p_y(\pi^+_{K_S^0})$ (MeV)',
'ks_pip_pz':r'$p_z(\pi^+_{K_S^0})$ (MeV)',
'ks_pim_px':r'$p_x(\pi^-_{K_S^0})$ (MeV)',
'ks_pim_py':r'$p_y(\pi^-_{K_S^0})$ (MeV)',
'ks_pim_pz':r'$p_z(\pi^-_{K_S^0})$ (MeV)',
'ks_E' :r'$E(K_S^0)$ (MeV)',
'ks_px' :r'$p_x(K_S^0)$ (MeV)',
'ks_py' :r'$p_y(K_S^0)$ (MeV)',
'ks_pz' :r'$p_z(K_S^0)$ (MeV)',
'ks_p' :r'$p(K_S^0)$ (MeV)',
'ks_m' :r'$m(K_S^0)$ (MeV)',
'ks_dpx' :r'Conservation $p_x(K_S^0)$ (MeV)',
'ks_dpy' :r'Conservation $p_y(K_S^0)$ (MeV)',
'ks_dpz' :r'Conservation $p_z(K_S^0)$ (MeV)',
'ks_dpe' :r'Conservation $E(K_S^0)$ (MeV)',
'phi_pip_px':r'$p_x(\pi^+_{\phi})$ (MeV)',
'phi_pip_py':r'$p_y(\pi^+_{\phi})$ (MeV)',
'phi_pip_pz':r'$p_z(\pi^+_{\phi})$ (MeV)',
'phi_pip_p' :r'$p(\pi^+_{\phi})$ (MeV)',
'phi_pim_px':r'$p_x(\pi^-_{\phi})$ (MeV)',
'phi_pim_py':r'$p_y(\pi^-_{\phi})$ (MeV)',
'phi_pim_pz':r'$p_z(\pi^-_{\phi})$ (MeV)',
'phi_pim_p' :r'$p(\pi^-_{\phi})$ (MeV)',
'phi_E' :r'$E(\phi)$ (MeV)',
'phi_px' :r'$p_x(\phi)$ (MeV)',
'phi_py' :r'$p_y(\phi)$ (MeV)',
'phi_pz' :r'$p_z(\phi)$ (MeV)',
'phi_p' :r'$p(\phi)$ (MeV)',
'phi_m' :r'$m(\phi)$ (MeV)',
'phi_dpx' :r'Conservation $p_x(\phi)$ (MeV)',
'phi_dpy' :r'Conservation $p_y(\phi)$ (MeV)',
'phi_dpz' :r'Conservation $p_z(\phi)$ (MeV)',
'phi_dpe' :r'Conservation $E(\phi)$ (MeV)',
'd0_E' :r'$E(D^0)$ (MeV)',
'd0_px' :r'$p_x(D^0)$ (MeV)',
'd0_py' :r'$p_y(D^0)$ (MeV)',
'd0_pz' :r'$p_z(D^0)$ (MeV)',
'd0_p' :r'$p(D^0)$ (MeV)',
'd0_m' :r'$m(D^0)$ (MeV)',
'd0_dpx' :r'Conservation $p_x(D^0)$ (MeV)',
'd0_dpy' :r'Conservation $p_y(D^0)$ (MeV)',
'd0_dpz' :r'Conservation $p_z(D^0)$ (MeV)',
'd0_dpe' :r'Conservation $E(D^0)$ (MeV)',}
for i in range(xi.shape[1]):
if filenames[i] not in ['ks_E']:
continue
m = (xi[:, i] - xi_gen[:, i])[cov[:, i, i] > 0]
cov2 = cov[cov[:, i, i] > 0]
m = m / cov2[:, i, i] ** 0.5
for _ in range(5):
mean, std = np.mean(m), np.std(m)
m = m[np.abs(m - mean) < 5.*std]
if np.std(m) < 0.000001:
return
x, bins, e = make_hist(m)
fig, ax = plt.subplots(figsize=(4, 3))
ax.errorbar(x, bins, e, linestyle='none', marker='.', markersize=4)
norm = (x[-1] - x[0]) / len(x) * sum(bins)
mu = 0
sigma = 1
x = np.linspace(mu - 3*sigma, mu + 3*sigma, 100)
plt.plot(x, norm * stats.norm.pdf(x, mu, sigma))
textstr = '\n'.join((
r"$\mu$ = {:0.3f}".format(mean),
r"$\sigma~$ = {:0.3f}".format(std)
))
props = dict(boxstyle='round', facecolor='wheat', alpha=0.5)
ax.text(0.05, 0.95, textstr, transform=ax.transAxes, fontsize=12,
verticalalignment='top', bbox=props)
ax.grid()
plt.xlabel('pull ' + labels[filenames[i]], fontsize=12)
plt.tight_layout()
plt.savefig('fig/d_meson/{}/pull_{}.pgf'.format(savedir, filenames[i]))
def plot_pull(xi, cov, pimgen, pipgen, savedir):
ks3 = pipgen + pimgen
xi0 = np.hstack((pipgen, pimgen, energy(
MASS_DICT['K0_S'], ks3).reshape(-1, 1), ks3))
filenames = ['pip_px', 'pip_py', 'pip_pz', 'pim_px', 'pim_py', 'pim_pz',
'ks_E', 'ks_px', 'ks_py', 'ks_pz']
labels = {'pip_px':r'$p_x(\pi+)$ (MeV)',
'pip_py':r'$p_y(\pi+)$ (MeV)',
'pip_pz':r'$p_z(\pi+)$ (MeV)',
'pim_px':r'$p_x(\pi-)$ (MeV)',
'pim_py':r'$p_y(\pi-)$ (MeV)',
'pim_pz':r'$p_z(\pi-)$ (MeV)',
'ks_E' :r'$E(K_S^0)$ (MeV)',
'ks_px' :r'$p_x(K_S^0)$ (MeV)',
'ks_py' :r'$p_y(K_S^0)$ (MeV)',
'ks_pz' :r'$p_z(K_S^0)$ (MeV)' }
for i in range(xi0.shape[1]):
fig, ax = plt.subplots(figsize=(4, 3))
# ax.figure()
m = (xi[:, i] - xi0[:, i])[cov[:, i, i] > 0]
cov2 = cov[cov[:, i, i] > 0]
m = m / cov2[:, i, i] ** 0.5
# m = (xi[:, i] - xi0[:, i]) / cov[:, i, i] ** 0.5
for _ in range(5):
mean, std = np.mean(m), np.std(m)
m = m[np.abs(m - mean) < 5.*std]
print(m.shape, np.mean(m), np.std(m))
if np.std(m) < 0.000001 or len(m) == 0:
return
x, bins, e = make_hist(m)
ax.errorbar(x, bins, e, linestyle='none', marker='.', markersize=4)
norm = (x[-1] - x[0]) / len(x) * sum(bins)
mu = 0
sigma = 1
x = np.linspace(mu - 3*sigma, mu + 3*sigma, 100)
ax.plot(x, norm * stats.norm.pdf(x, mu, sigma))
# textstr = '\n'.join((
# r"$\sigma_{{after}}$ = {:0.3f}".format(fit_std),
# r"$\sigma_{{before}}$ = {:0.3f}".format(unfit_std),
# r"$\mu_{{after}}$ = {:0.3f}".format(fit_mean),
# r"$\mu_{{before}}$ = {:0.3f}".format(unfit_mean)
# ))
# props = dict(boxstyle='round', facecolor='wheat', alpha=0.5)
# ax.text(0.05, 0.95, textstr, transform=ax.transAxes, fontsize=12,
# verticalalignment='top', bbox=props)
ax.plot([], [], ' ', label=r"$mean$ {:0.3f}".format(mean))
ax.plot([], [], ' ', label=r"$\sigma~$ {:0.3f}".format(std))
ax.legend()
ax.grid()
ax.xlabel(labels[filenames[i]], fontsize=12)
ax.tight_layout()
ax.show()
def plot_params(xi, xi0, pimgen, pipgen, savedir):
ks3 = pipgen + pimgen
xi_gen = np.hstack((pipgen, pimgen,
(energy(MASS_DICT['pi+'], pipgen) + energy(MASS_DICT['pi+'], pimgen)).reshape(-1, 1),
ks3))
def augment(xi):
ks_m = np.sqrt(xi[:, 6]**2 - np.sum(xi[:, 7:10]
** 2, axis=-1)).reshape(-1, 1)
ks_p = np.sqrt(np.sum(xi[:, 7:10]**2, axis=-1)).reshape(-1, 1)
pip_p = np.sqrt(np.sum(xi[:, :3]**2, axis=-1)).reshape(-1, 1)
pim_p = np.sqrt(np.sum(xi[:, 3:6]**2, axis=-1)).reshape(-1, 1)
dpx = (xi[:, 0] + xi[:, 3] - xi[:, 7]).reshape(-1, 1)
dpy = (xi[:, 1] + xi[:, 4] - xi[:, 8]).reshape(-1, 1)
dpz = (xi[:, 2] + xi[:, 5] - xi[:, 9]).reshape(-1, 1)
dpe = (energy(MASS_DICT['pi+'], xi[:, 0:3]) + energy(MASS_DICT['pi+'], xi[:, 3:6]) -
xi[:, 6]).reshape(-1, 1)
return np.hstack((xi[:, :10], ks_m, ks_p, pip_p, pim_p, dpx, dpy, dpz, dpe))
xi = augment(xi)
xi0 = augment(xi0)
xi_gen = augment(xi_gen)
filenames = ['pip_px', 'pip_py', 'pip_pz', 'pim_px', 'pim_py', 'pim_pz',
'ks_E', 'ks_px', 'ks_py', 'ks_pz', 'ks_m', 'ks_p', 'pip_p',
'pim_p', 'dpx', 'dpy', 'dpz', 'dpe']
labels = { 'pip_px':r'$p_x(\pi+)$ (MeV)',
'pip_py':r'$p_y(\pi+)$ (MeV)',
'pip_pz':r'$p_z(\pi+)$ (MeV)',
'pim_px':r'$p_x(\pi-)$ (MeV)',
'pim_py':r'$p_y(\pi-)$ (MeV)',
'pim_pz':r'$p_z(\pi-)$ (MeV)',
'ks_E' :r'$E(K_S^0)$ (MeV)',
'ks_px' :r'$p_x(K_S^0)$ (MeV)',
'ks_py' :r'$p_y(K_S^0)$ (MeV)',
'ks_pz' :r'$p_z(K_S^0)$ (MeV)',
'ks_m' :r'$m(K_S^0)$ (MeV)',
'ks_p' :r'$p(K_S^0)$ (MeV)',
'pip_p' :r'$p(\pi+)$ (MeV)',
'pim_p' :r'$p(\pi-)$ (MeV)',
'dpx' :r'Conservation $p_x$ (MeV)',
'dpy' :r'Conservation $p_y$ (MeV)',
'dpz' :r'Conservation $p_z$ (MeV)',
'dpe' :r'Conservation $E$ (MeV)',}
for i in range(xi.shape[1]):
fitted = xi[:, i] - xi_gen[:, i]
unfitted = xi0[:, i] - xi_gen[:, i]
for _ in range(5):
fit_mean, fit_std = np.mean(fitted), np.std(fitted)
fitted = fitted[np.abs(fitted - fit_mean) < 5.*fit_std]
unfit_mean, unfit_std = np.mean(unfitted), np.std(unfitted)
unfitted = unfitted[np.abs(unfitted - unfit_mean) < 5.*unfit_std]
plt.figure(figsize=(4, 3))
if len(fitted) != 0:
plt.errorbar(*make_hist(fitted, density=True),
linestyle='none', marker='.', markersize=4, label='fit')
if len(unfitted) != 0:
plt.errorbar(*make_hist(unfitted, density=True),
linestyle='none', marker='.', markersize=4, label='unfit')
plt.plot([], [], ' ', label=r"$\sigma_{{fit}}$ {:0.3f}".format(fit_std))
plt.plot([], [], ' ', label=r"$\sigma_{{unfit}}$ {:0.3f}".format(unfit_std))
plt.plot([], [], ' ', label=r"$\mu_{{fit}}$ {:0.3f}".format(fit_mean))
plt.plot([], [], ' ', label=r"$\mu_{{unfit}}$ {:0.3f}".format(unfit_mean))
plt.legend(loc='upper right')
plt.grid()
plt.xlabel(labels[filenames[i]], fontsize=14)
plt.tight_layout()
pathlib.Path('fig/kaon/{}'.format(savedir)).mkdir(parents=True, exist_ok=True)
plt.savefig('fig/kaon/{}/fit_{}.pgf'.format(savedir, filenames[i]))
def plot_params_d0(xi, xi0, p3_ks_pip_gen, p3_ks_pim_gen, p3_phi_pip_gen, p3_phi_pim_gen, savedir):
ks3 = p3_ks_pip_gen + p3_ks_pim_gen
phi3 = p3_phi_pip_gen + p3_phi_pim_gen
d03 = ks3 + phi3
eks_gen = energy(MASS_DICT['K0_S'], ks3).reshape(-1, 1)
ephi_gen = energy(MASS_DICT['phi'], phi3).reshape(-1, 1)
ed0_gen = energy(MASS_DICT['D0'], d03).reshape(-1, 1)
xi_gen = np.hstack((p3_ks_pip_gen, p3_ks_pim_gen, eks_gen, ks3,
p3_phi_pip_gen, p3_phi_pim_gen, ephi_gen, phi3,
ed0_gen, d03))
if (savedir.split('/')[0] == 'reffit'):
xi = np.hstack((xi[:, 0:10], xi[:,15:25], xi[:,30:34]))
xi0 = np.hstack((xi0[:, 0:10], xi0[:,15:25], xi0[:,30:34]))
def augment(xi):
ks_pip_p = np.sqrt(np.sum(xi[:, 0:3]**2, axis=-1)).reshape(-1, 1)
ks_pim_p = np.sqrt(np.sum(xi[:, 3:6]**2, axis=-1)).reshape(-1, 1)
ks_p = np.sqrt(np.sum(xi[:, 7:10]**2, axis=-1)).reshape(-1, 1)
ks_m = np.sqrt(xi[:, 6]**2 - np.sum(xi[:, 7:10]** 2, axis=-1)).reshape(-1, 1)
phi_pip_p = np.sqrt(np.sum(xi[:, 10:13]**2, axis=-1)).reshape(-1, 1)
phi_pim_p = np.sqrt(np.sum(xi[:, 13:16]**2, axis=-1)).reshape(-1, 1)
phi_p = np.sqrt(np.sum(xi[:, 17:20]**2, axis=-1)).reshape(-1, 1)
phi_m = np.sqrt(xi[:, 16]**2 - np.sum(xi[:, 17:20]** 2, axis=-1)).reshape(-1, 1)
d0_p = np.sqrt(np.sum(xi[:, 21:24]**2, axis=-1)).reshape(-1, 1)
d0_m = np.sqrt(xi[:, 20]**2 - np.sum(xi[:, 21:24]** 2, axis=-1)).reshape(-1, 1)
ks_dpx = (xi[:, 0] + xi[:, 3] - xi[:, 7]).reshape(-1, 1)
ks_dpy = (xi[:, 1] + xi[:, 4] - xi[:, 8]).reshape(-1, 1)
ks_dpz = (xi[:, 2] + xi[:, 5] - xi[:, 9]).reshape(-1, 1)
ks_dpe = (energy(MASS_DICT['pi+'], xi[:, 0:3]) + energy(MASS_DICT['pi+'], xi[:, 3:6]) - xi[:, 6]).reshape(-1, 1)
phi_dpx = (xi[:, 10] + xi[:, 13] - xi[:, 17]).reshape(-1, 1)
phi_dpy = (xi[:, 11] + xi[:, 14] - xi[:, 18]).reshape(-1, 1)
phi_dpz = (xi[:, 12] + xi[:, 15] - xi[:, 19]).reshape(-1, 1)
phi_dpe = (energy(MASS_DICT['pi+'], xi[:, 10:13]) + energy(MASS_DICT['pi+'], xi[:, 13:16]) - xi[:, 16]).reshape(-1, 1)
d0_dpx = (xi[:, 7] + xi[:, 17] - xi[:, 21]).reshape(-1, 1)
d0_dpy = (xi[:, 8] + xi[:, 18] - xi[:, 22]).reshape(-1, 1)
d0_dpz = (xi[:, 9] + xi[:, 19] - xi[:, 23]).reshape(-1, 1)
d0_dpe = (xi[:, 6] + xi[:, 16] - xi[:, 20]).reshape(-1, 1)
return np.hstack((xi, ks_pip_p, ks_pim_p, ks_p, ks_m, phi_pip_p, phi_pim_p, phi_p, phi_m, d0_p, d0_m,
ks_dpx, ks_dpy, ks_dpz, ks_dpe, phi_dpx, phi_dpy, phi_dpz, phi_dpe, d0_dpx, d0_dpy, d0_dpz, d0_dpe))
xi = augment(xi)
xi0 = augment(xi0)
xi_gen = augment(xi_gen)
filenames = ['ks_pip_px', 'ks_pip_py', 'ks_pip_pz', 'ks_pim_px', 'ks_pim_py', 'ks_pim_pz', 'ks_E', 'ks_px', 'ks_py', 'ks_pz',
'phi_pip_px', 'phi_pip_py', 'phi_pip_pz', 'phi_pim_px', 'phi_pim_py', 'phi_pim_pz', 'phi_E', 'phi_px', 'phi_py', 'phi_pz',
'd0_E', 'd0_px', 'd0_py', 'd0_pz',
'ks_pip_p', 'ks_pim_p', 'ks_p', 'ks_m',
'phi_pip_p', 'phi_pim_p', 'phi_p', 'phi_m',
'd0_p', 'd0_m',
'ks_dpx', 'ks_dpy', 'ks_dpz', 'ks_dpe',
'phi_dpx', 'phi_dpy', 'phi_dpz', 'phi_dpe',
'd0_dpx', 'd0_dpy', 'd0_dpz', 'd0_dpe',
]
labels = { 'ks_pip_px':r'$p_x(\pi^+_{K_S^0})$ (MeV)',
'ks_pip_py':r'$p_y(\pi^+_{K_S^0})$ (MeV)',
'ks_pip_pz':r'$p_z(\pi^+_{K_S^0})$ (MeV)',
'ks_pip_p' :r'$p(\pi^+_{K_S^0})$ (MeV)',
'ks_pim_px':r'$p_x(\pi^-_{K_S^0})$ (MeV)',
'ks_pim_py':r'$p_y(\pi^-_{K_S^0})$ (MeV)',
'ks_pim_pz':r'$p_z(\pi^-_{K_S^0})$ (MeV)',
'ks_pim_p' :r'$p(\pi^-_{K_S^0})$ (MeV)',
'ks_E' :r'$E(K_S^0)$ (MeV)',
'ks_px' :r'$p_x(K_S^0)$ (MeV)',
'ks_py' :r'$p_y(K_S^0)$ (MeV)',
'ks_pz' :r'$p_z(K_S^0)$ (MeV)',
'ks_p' :r'$p(K_S^0)$ (MeV)',
'ks_m' :r'$m(K_S^0)$ (MeV)',
'ks_dpx' :r'Conservation $p_x(K_S^0)$ (MeV)',
'ks_dpy' :r'Conservation $p_y(K_S^0)$ (MeV)',
'ks_dpz' :r'Conservation $p_z(K_S^0)$ (MeV)',
'ks_dpe' :r'Conservation $E(K_S^0)$ (MeV)',
'phi_pip_px':r'$p_x(\pi^+_{\phi})$ (MeV)',
'phi_pip_py':r'$p_y(\pi^+_{\phi})$ (MeV)',
'phi_pip_pz':r'$p_z(\pi^+_{\phi})$ (MeV)',
'phi_pip_p' :r'$p(\pi^+_{\phi})$ (MeV)',
'phi_pim_px':r'$p_x(\pi^-_{\phi})$ (MeV)',
'phi_pim_py':r'$p_y(\pi^-_{\phi})$ (MeV)',
'phi_pim_pz':r'$p_z(\pi^-_{\phi})$ (MeV)',
'phi_pim_p' :r'$p(\pi^-_{\phi})$ (MeV)',
'phi_E' :r'$E(\phi)$ (MeV)',
'phi_px' :r'$p_x(\phi)$ (MeV)',
'phi_py' :r'$p_y(\phi)$ (MeV)',
'phi_pz' :r'$p_z(\phi)$ (MeV)',
'phi_p' :r'$p(\phi)$ (MeV)',
'phi_m' :r'$m(\phi)$ (MeV)',
'phi_dpx' :r'Conservation $p_x(\phi)$ (MeV)',
'phi_dpy' :r'Conservation $p_y(\phi)$ (MeV)',
'phi_dpz' :r'Conservation $p_z(\phi)$ (MeV)',
'phi_dpe' :r'Conservation $E(\phi)$ (MeV)',
'd0_E' :r'$p_x(D^0)$ (MeV)',
'd0_px' :r'$p_x(D^0)$ (MeV)',
'd0_py' :r'$p_y(D^0)$ (MeV)',
'd0_pz' :r'$p_z(D^0$ (MeV)',
'd0_p' :r'$p(D^0)$ (MeV)',
'd0_m' :r'$m(D^0)$ (MeV)',
'd0_dpx' :r'Conservation $p_x(D^0)$ (MeV)',
'd0_dpy' :r'Conservation $p_y(D^0)$ (MeV)',
'd0_dpz' :r'Conservation $p_z(D^0)$ (MeV)',
'd0_dpe' :r'Conservation $E(D^0)$ (MeV)',}
for i in range(xi.shape[1]):
if filenames[i] not in ['ks_dpe']:
continue
fitted = xi[:, i] - xi_gen[:, i]
unfitted = xi0[:, i] - xi_gen[:, i]
for _ in range(5):
fit_mean, fit_std = np.mean(fitted), np.std(fitted)
fitted = fitted[np.abs(fitted - fit_mean) < 5.*fit_std]
unfit_mean, unfit_std = np.mean(unfitted), np.std(unfitted)
unfitted = unfitted[np.abs(unfitted - unfit_mean) < 5.*unfit_std]
fig, ax = plt.subplots(figsize=(4, 3))
# plt.figure(figsize=(4,3))
# ax.figure()
if len(fitted) != 0:
plt.errorbar(*make_hist(fitted, density=True),
linestyle='none', marker='.', markersize=4, label='after fit')
if 1e-1 > (fit_std / unfit_std) or (fit_std / unfit_std) > 1e1:
plt.legend(loc='upper right')
plt.grid()
plt.xlabel(labels[filenames[i]], fontsize=16)
# textstr = '\n'.join((
# r"$\sigma_{{after}}$ = {:0.3f}".format(fit_std),
# r"$\mu_{{after}}$ = {:0.3f}".format(fit_mean),
# ))
# props = dict(boxstyle='round', facecolor='wheat', alpha=0.5)
# ax.text(0.05, 0.95, textstr, transform=ax.transAxes, fontsize=12,
# verticalalignment='top', bbox=props)
# ax.legend(loc='upper right', fontsize=12)
# ax.grid()
plt.xlabel(labels[filenames[i]], fontsize=12)
plt.tight_layout()
pathlib.Path('fig/d_meson/{}'.format(savedir)).mkdir(parents=True, exist_ok=True)
plt.savefig('fig/d_meson/{}/fit_{}.pgf'.format(savedir, filenames[i]))
def plot_evlolution_params_d0(xi_full, p3_ks_pip_gen, p3_ks_pim_gen, p3_phi_pip_gen, p3_phi_pim_gen, savedir):
iter = -1
for xi in xi_full:
iter += 1
ks3 = p3_ks_pip_gen + p3_ks_pim_gen
phi3 = p3_phi_pip_gen + p3_phi_pim_gen
d03 = ks3 + phi3
eks_gen = energy(MASS_DICT['K0_S'], ks3).reshape(-1, 1)
ephi_gen = energy(MASS_DICT['phi'], phi3).reshape(-1, 1)
ed0_gen = energy(MASS_DICT['D0'], d03).reshape(-1, 1)
xi_gen = np.hstack((p3_ks_pip_gen, p3_ks_pim_gen, eks_gen, ks3,
p3_phi_pip_gen, p3_phi_pim_gen, ephi_gen, phi3,
ed0_gen, d03))
if (savedir.split('/')[0] == 'reffit'):
xi = np.hstack((xi[:, 0:10], xi[:,15:25], xi[:,30:34]))
# xi0 = np.hstack((xi0[:, 0:10], xi0[:,15:25], xi0[:,30:34]))
def augment(xi):
ks_pip_p = np.sqrt(np.sum(xi[:, 0:3]**2, axis=-1)).reshape(-1, 1)
ks_pim_p = np.sqrt(np.sum(xi[:, 3:6]**2, axis=-1)).reshape(-1, 1)
ks_p = np.sqrt(np.sum(xi[:, 7:10]**2, axis=-1)).reshape(-1, 1)
ks_m = np.sqrt(xi[:, 6]**2 - np.sum(xi[:, 7:10]** 2, axis=-1)).reshape(-1, 1)
phi_pip_p = np.sqrt(np.sum(xi[:, 10:13]**2, axis=-1)).reshape(-1, 1)
phi_pim_p = np.sqrt(np.sum(xi[:, 13:16]**2, axis=-1)).reshape(-1, 1)
phi_p = np.sqrt(np.sum(xi[:, 17:20]**2, axis=-1)).reshape(-1, 1)
phi_m = np.sqrt(xi[:, 16]**2 - np.sum(xi[:, 17:20]** 2, axis=-1)).reshape(-1, 1)
d0_p = np.sqrt(np.sum(xi[:, 21:24]**2, axis=-1)).reshape(-1, 1)
d0_m = np.sqrt(xi[:, 20]**2 - np.sum(xi[:, 21:24]** 2, axis=-1)).reshape(-1, 1)
ks_dpx = (xi[:, 0] + xi[:, 3] - xi[:, 7]).reshape(-1, 1)
ks_dpy = (xi[:, 1] + xi[:, 4] - xi[:, 8]).reshape(-1, 1)
ks_dpz = (xi[:, 2] + xi[:, 5] - xi[:, 9]).reshape(-1, 1)
ks_dpe = (energy(MASS_DICT['pi+'], xi[:, 0:3]) + energy(MASS_DICT['pi+'], xi[:, 3:6]) - xi[:, 6]).reshape(-1, 1)
phi_dpx = (xi[:, 10] + xi[:, 13] - xi[:, 17]).reshape(-1, 1)
phi_dpy = (xi[:, 11] + xi[:, 14] - xi[:, 18]).reshape(-1, 1)
phi_dpz = (xi[:, 12] + xi[:, 15] - xi[:, 19]).reshape(-1, 1)
phi_dpe = (energy(MASS_DICT['pi+'], xi[:, 10:13]) + energy(MASS_DICT['pi+'], xi[:, 13:16]) - xi[:, 16]).reshape(-1, 1)
d0_dpx = (xi[:, 7] + xi[:, 17] - xi[:, 21]).reshape(-1, 1)
d0_dpy = (xi[:, 8] + xi[:, 18] - xi[:, 22]).reshape(-1, 1)
d0_dpz = (xi[:, 9] + xi[:, 19] - xi[:, 23]).reshape(-1, 1)
d0_dpe = (xi[:, 6] + xi[:, 16] - xi[:, 20]).reshape(-1, 1)
return np.hstack((xi, ks_pip_p, ks_pim_p, ks_p, ks_m, phi_pip_p, phi_pim_p, phi_p, phi_m, d0_p, d0_m,
ks_dpx, ks_dpy, ks_dpz, ks_dpe, phi_dpx, phi_dpy, phi_dpz, phi_dpe, d0_dpx, d0_dpy, d0_dpz, d0_dpe))
xi = augment(xi)
# xi0 = augment(xi0)
xi_gen = augment(xi_gen)
filenames = ['ks_pip_px', 'ks_pip_py', 'ks_pip_pz', 'ks_pim_px', 'ks_pim_py', 'ks_pim_pz', 'ks_E', 'ks_px', 'ks_py', 'ks_pz',
'phi_pip_px', 'phi_pip_py', 'phi_pip_pz', 'phi_pim_px', 'phi_pim_py', 'phi_pim_pz', 'phi_E', 'phi_px', 'phi_py', 'phi_pz',
'd0_E', 'd0_px', 'd0_py', 'd0_pz',
'ks_pip_p', 'ks_pim_p', 'ks_p', 'ks_m',
'phi_pip_p', 'phi_pim_p', 'phi_p', 'phi_m',
'd0_p', 'd0_m',
'ks_dpx', 'ks_dpy', 'ks_dpz', 'ks_dpe',
'phi_dpx', 'phi_dpy', 'phi_dpz', 'phi_dpe',
'd0_dpx', 'd0_dpy', 'd0_dpz', 'd0_dpe',
]
labels = { 'ks_pip_px':r'$p_x(\pi^+_{K_S^0})$ (MeV)',
'ks_pip_py':r'$p_y(\pi^+_{K_S^0})$ (MeV)',
'ks_pip_pz':r'$p_z(\pi^+_{K_S^0})$ (MeV)',
'ks_pip_p' :r'$p(\pi^+_{K_S^0})$ (MeV)',
'ks_pim_px':r'$p_x(\pi^-_{K_S^0})$ (MeV)',
'ks_pim_py':r'$p_y(\pi^-_{K_S^0})$ (MeV)',
'ks_pim_pz':r'$p_z(\pi^-_{K_S^0})$ (MeV)',
'ks_pim_p' :r'$p(\pi^-_{K_S^0})$ (MeV)',
'ks_E' :r'$E(K_S^0)$ (MeV)',
'ks_px' :r'$p_x(K_S^0)$ (MeV)',
'ks_py' :r'$p_y(K_S^0)$ (MeV)',
'ks_pz' :r'$p_z(K_S^0)$ (MeV)',
'ks_p' :r'$p(K_S^0)$ (MeV)',
'ks_m' :r'$m(K_S^0)$ (MeV)',
'ks_dpx' :r'Conservation $p_x(K_S^0)$ (MeV)',
'ks_dpy' :r'Conservation $p_y(K_S^0)$ (MeV)',
'ks_dpz' :r'Conservation $p_z(K_S^0)$ (MeV)',
'ks_dpe' :r'Conservation $E(K_S^0)$ (MeV)',
'phi_pip_px':r'$p_x(\pi^+_{\phi})$ (MeV)',
'phi_pip_py':r'$p_y(\pi^+_{\phi})$ (MeV)',
'phi_pip_pz':r'$p_z(\pi^+_{\phi})$ (MeV)',
'phi_pip_p' :r'$p(\pi^+_{\phi})$ (MeV)',
'phi_pim_px':r'$p_x(\pi^-_{\phi})$ (MeV)',
'phi_pim_py':r'$p_y(\pi^-_{\phi})$ (MeV)',
'phi_pim_pz':r'$p_z(\pi^-_{\phi})$ (MeV)',
'phi_pim_p' :r'$p(\pi^-_{\phi})$ (MeV)',
'phi_E' :r'$E(\phi)$ (MeV)',
'phi_px' :r'$p_x(\phi)$ (MeV)',
'phi_py' :r'$p_y(\phi)$ (MeV)',
'phi_pz' :r'$p_z(\phi)$ (MeV)',
'phi_p' :r'$p(\phi)$ (MeV)',
'phi_m' :r'$m(\phi)$ (MeV)',
'phi_dpx' :r'Conservation $p_x(\phi)$ (MeV)',
'phi_dpy' :r'Conservation $p_y(\phi)$ (MeV)',
'phi_dpz' :r'Conservation $p_z(\phi)$ (MeV)',
'phi_dpe' :r'Conservation $E(\phi)$ (MeV)',
'd0_E' :r'$p_x(D^0)$ (MeV)',
'd0_px' :r'$p_x(D^0)$ (MeV)',
'd0_py' :r'$p_y(D^0)$ (MeV)',
'd0_pz' :r'$p_z(D^0$ (MeV)',
'd0_p' :r'$p(D^0)$ (MeV)',
'd0_m' :r'$m(D^0)$ (MeV)',
'd0_dpx' :r'Conservation $p_x(D^0)$ (MeV)',
'd0_dpy' :r'Conservation $p_y(D^0)$ (MeV)',
'd0_dpz' :r'Conservation $p_z(D^0)$ (MeV)',
'd0_dpe' :r'Conservation $E(D^0)$ (MeV)',}
for i in [33,]:#range(xi.shape[1]):
fitted = xi[:, i] - xi_gen[:, i]
# unfitted = xi0[:, i] - xi_gen[:, i]
for _ in range(5):
fit_mean, fit_std = np.mean(fitted), np.std(fitted)
fitted = fitted[np.abs(fitted - fit_mean) < 5.*fit_std]
# unfit_mean, unfit_std = np.mean(unfitted), np.std(unfitted)
# unfitted = unfitted[np.abs(unfitted - unfit_mean) < 5.*unfit_std]
# plt.figure(figsize=(8, 6))
# if 1e-1 > (fit_std / unfit_std) or (fit_std / unfit_std) > 1e1:
# plt.subplot(1, 2, 1)
if len(fitted) != 0:
if fit_std > 1e-3:
plt.errorbar(*make_hist(fitted, range=[-0.1, 0.1], density=True),
linestyle='none', marker='.', markersize=4, label='fit')
# if 1e-1 > (fit_std / unfit_std) or (fit_std / unfit_std) > 1e1:
plt.plot([], [], ' ', label=r"$\sigma_{{fit}}$ {:0.3f}".format(fit_std))
# plt.plot([], [], ' ', label=r"$\sigma_{{unfit}}$ {:0.3f}".format(unfit_std))
plt.plot([], [], ' ', label=r"$\mu_{{fit}}$ {:0.3f}".format(fit_mean))
# plt.plot([], [], ' ', label=r"$\mu_{{unfit}}$ {:0.3f}".format(unfit_mean))
plt.legend(loc='upper right')
plt.grid()
plt.xlabel(labels[filenames[i]], fontsize=16)
# plt.subplot(1, 2, 2)
# if len(unfitted) != 0:
# plt.errorbar(*make_hist(unfitted, density=True),
# linestyle='none', marker='.', markersize=4, label='unfit')
# plt.plot([], [], ' ', label=r"$\sigma_{{fit}}$ {:0.3f}".format(fit_std))
# plt.plot([], [], ' ', label=r"$\sigma_{{unfit}}$ {:0.3f}".format(unfit_std))
# plt.plot([], [], ' ', label=r"$\mu_{{fit}}$ {:0.3f}".format(fit_mean))
# plt.plot([], [], ' ', label=r"$\mu_{{unfit}}$ {:0.3f}".format(unfit_mean))
plt.legend(loc='upper right')
plt.grid()
plt.xlabel("Evolution: {} {}, iter: {}".format(labels[filenames[i]], savedir, iter), fontsize=16)
plt.tight_layout()
# pathlib.Path('fig/d_meson/{}'.format(savedir)).mkdir(parents=True, exist_ok=True)
# plt.savefig('fig/d_meson/{}/evolution_{}.png'.format(savedir, filenames[i]))
plt.show()