def DEC_evol(steps = [0,6,9,19]):
"""
Draw evolution of DECs with respect to plastic deformation
"""
from MP.mat import mech # mech is a module
FlowCurve = mech.FlowCurve
from mst_ex import use_intp_sfig, return_vf
from rs import ResidualStress, filter_psi3
model_rs = ResidualStress(
mod_ext=None,fnmod_ig='igstrain_fbulk_ph1.out',
fnmod_sf='igstrain_fbulk_ph1.out',i_ip=1)
fc = FlowCurve(name='Weighted Average')
fc.get_model(fn='STR_STR.OUT')
fc.get_eqv()
phi = model_rs.dat_model.phi
nphi = model_rs.dat_model.nphi
psi = model_rs.dat_model.psi
sin2psi = sin2psi_opt(psi,1)
DEC_raw = model_rs.dat_model.sf.copy()
vf_raw, dum = return_vf()
DEC_raw, psi, sin2psi, vf_raw = filter_psi3(
sin2psi.copy(), [-0.5,0.5],
DEC_raw, psi.copy(), sin2psi.copy(), vf_raw.copy())
## filter DEC_raw
DEC_raw[DEC_raw==0]=np.nan
evm = fc.epsilon_vm.copy()
nstp = fc.nstp
fig = plt.figure()
# steps = [0,3,5,9]
nstp = len(steps)
gs=gridspec.GridSpec(
nstp,nphi,
wspace=0,hspace=0,left=0.25,right=0.8,top=0.8)
axes=[]; axev=[]; axs=[]; axv=[]
for i in range(len(steps)):
axes.append([])
axev.append([])
for j in range(nphi):
ax = fig.add_subplot(gs[i,j])
av = ax.twinx()
axs.append(ax);axes[i].append(ax); axv.append(av); axev[i].append(av)
ax.locator_params(nbins=4)
lab1 = r'$\mathbb{F}_{11}$'; lab2 = r'$\mathbb{F}_{22}$'
ax.plot(sin2psi, DEC_raw[steps[i],0,j,:]*1e6,'k-',label=lab1)
ax.plot(sin2psi, DEC_raw[steps[i],1,j,:]*1e6,'-',color='gray',label=lab2)
av.plot(sin2psi, vf_raw[steps[i],j,:],color='gray',linewidth=3,alpha=0.7)
ax.plot(0,0,'-',color='gray',linewidth=3,alpha=0.7,label='Vol. F')
if i==len(steps)-1 and j==0: pass
else:
mpl_lib.rm_lab(ax,axis='x')
mpl_lib.rm_lab(ax,axis='y')
if i==len(steps)-1 and j==nphi-1: pass
else:
mpl_lib.rm_lab(av,axis='x')
mpl_lib.rm_lab(av,axis='y')
tune_xy_lim(axs)
tune_x_lim(axs,axis='y')
tune_xy_lim(axv)
tune_x_lim(axv,axis='y')
for i in range(len(steps)):#range(nstp):
if i ==0: s = r'$\bar{E}^{VM}=%.2f$'%fc.epsilon_vm[steps[i]]
else: s = r'$%.2f$'%fc.epsilon_vm[steps[i]]
axes[i][-1].annotate(
s=s,
verticalalignment='center',
horizontalalignment='center',
rotation=270,
size=10,xy=(1.60,0.5),
xycoords='axes fraction')
for j in range(nphi):
s = r'$\phi=%.1f^\circ{}$'%(phi[j]*180./np.pi)
axes[0][j].annotate(
s=s,
horizontalalignment='center',
size=10,xy=(0.5,1.2),
xycoords='axes fraction')
fancy_legend(axes[0][0],size=10,nscat=1,ncol=1,
bbox_to_anchor=(-0.1,1))
deco(ax=axes[-1][0],ft=10,iopt=1,hkl='211',ipsi_opt=1)
deco(ax=axev[-1][-1],ft=10,iopt=7,hkl='211',ipsi_opt=1)
axev[-1][-1].grid('off');
for i in range(len(axv)):
axv[i].set_ylim(0.,0.3)
axv[i].locator_params(nbins=4)
fig.savefig("dec_evol.pdf")
fig.savefig("dec_evol.png")
def DEC_intp(ss=[1,2,4],intps=[0,3,4],inds=[79,90,120]):
"""
DEC interpolation comparison
intps=0: NN (piecewise linear interpolation)
=1: Assign nearest data
=2: Cubic
=3: Quadratic
=4: Linear fit
=5: Poly2
=6: poly3
=7: Place-holder for power law fit
=8: zero
=9: slinear
inds=[70,40]
"""
from MP.mat import mech # mech is a module
FlowCurve = mech.FlowCurve
from mst_ex import use_intp_sfig, return_vf
from rs import ResidualStress
model_rs = ResidualStress(
mod_ext=None,fnmod_ig='igstrain_fbulk_ph1.out',
fnmod_sf='igstrain_fbulk_ph1.out',i_ip=1)
psi = model_rs.dat_model.psi
sin2psi = sin2psi_opt(psi,1)
DEC_raw = model_rs.dat_model.sf.copy()
fc = FlowCurve(name='Weighted Average')
fc.get_model(fn='STR_STR.OUT')
fc.get_eqv()
evm = fc.epsilon_vm.copy()
fig = plt.figure()
gs=gridspec.GridSpec(
len(ss),len(intps),
wspace=0,hspace=0,left=0.25,right=0.8,top=0.8)
axes=[]
ms=['o','x','+','d','>','t']
for i in range(len(ss)):
axes.append([])
for j in range(len(intps)):
## interpolated DECs
print i,j
print ss[i],intps[j]
# if i==1 and j==0:
# raise IOError
_sf_, _ig_ = use_intp_sfig(
ss[i],iopt=intps[j],
iplot=False,iwgt=False)
dum = _sf_.sf.swapaxes(-1,-2).swapaxes(-2,-3)
DEC_intp = dum.copy()
ax = fig.add_subplot(gs[i,j])
ax.locator_params(nbins=4)
axes[i].append(ax)
for k in range(len(inds)):
ind = inds[k]
val_sin2psi = sin2psi[ind]
val_psi = psi[ind] * 180./np.pi
lab1, lab2 = None, None
if k==0:
lab1=r'Actual $\mathbb{F}_{ij}$'
lab2=r'Interpolated $\mathbb{F}^{\ I}_{ij}$'
y_raw = DEC_raw[:,0,0,ind]*1e6
ax.plot(evm, y_raw,'k-',
label=lab1)
y_intp = DEC_intp[:,0,0,ind]*1e12
ax.plot(evm,y_intp,'k--',
label=lab2)
ax.fill_between(evm,y_raw,y_intp,
facecolor='gray',alpha=0.5)
ax.plot(
evm[::ss[i]],
DEC_raw[:,0,0,ind][::ss[i]]*1e6,
ms[k],mfc='None',mec='black',
label=r'$\psi=%2.0f^\circ{}$'%val_psi)
if i==len(ss)-1 and j==0:
pass
else:
mpl_lib.rm_lab(ax,axis='x')
mpl_lib.rm_lab(ax,axis='y')
deco(ax=axes[-1][0],ft=10,iopt=6,ij=[1,1])
axes[-1][0].grid('off')
tune_xy_lim(fig.axes)
tune_x_lim(fig.axes,axis='y')
## annotations
for j in range(len(intps)):
if j==0: s = 'Piecewise'
if j==1: s = 'Quadratic'
if j==2: s = 'Linear fit'
if j==3: s = 'Power law fit'
axes[0][j].annotate(
s=s, horizontalalignment='center',
size=10,xy=(0.5,1.2),
xycoords='axes fraction')
for i in range(len(ss)):
s = r'$f^{\ \mathbb{F}}=^1/_{%i}$'%ss[i]
axes[i][-1].annotate(
s=s,
verticalalignment='center',
horizontalalignment='center',
rotation=270,
size=10,xy=(1.20,0.5),
xycoords='axes fraction')
fancy_legend(axes[0][0],size=10,nscat=1,ncol=1,
bbox_to_anchor=(-0.1,1))
fig.savefig('dec_intp.pdf')
fig.savefig('dec_intp.png')
return model_rs
def tabular_figs02():
fig=plt.figure()
axes=[]
for i in range(len(sigmas)):
axes.append([])
for j in range(len(nbins)):
if sigmas[i]==0: iscatter=False
else: iscatter= True
rst = main(
## main variables
sigma=sigmas[i], psi_nbin=nbins[j],
## minor
sin2psimx=0.5, iscatter=iscatter,
iplot=False,dec_inv_frq=1,
dec_interp=1)
model_rs, flow_weight, flow_dsa = rst
ax=fig.add_subplot(gs[i,j])
ax.locator_params(nbins=4)
axes[i].append(ax)
# von mises
lab1=r'$(\langle \sigma^c \rangle)^{VM}$'
lab2=r'$(\sigma^\mathrm{RS})^{VM}$'
ax.plot(flow_weight.epsilon_vm,
flow_weight.sigma_vm,'k-',label=lab1)
ax.plot(flow_dsa.epsilon_vm,
flow_dsa.sigma_vm,'kx',label=lab2)
if i==len(sigmas)-1 and j==0:
#VM Deco
axes_label.__vm__(ax=ax,ft=10)
else:
mpl_lib.rm_lab(ax,axis='x')
mpl_lib.rm_lab(ax,axis='y')
for iax in range(len(fig.axes)):
fig.axes[iax].set_ylim(0,)
fig.axes[iax].set_xlim(0,)
tune_xy_lim(fig.axes)
tune_x_lim(fig.axes,axis='y')
## annotations
for j in range(len(nbins)):
if j==0: s=r'$N^\psi$=%i'%nbins[j]
else:
s = '%i'%nbins[j]
axes[0][j].annotate(
s=s,horizontalalignment='center',
size=14,xy=(0.5,1.2),
xycoords='axes fraction')
for i in range(len(sigmas)):
if i==0:
s=r'$s^\mathrm{CSE}$=%i'%(
sigmas[i]*(10**6))
elif i==len(sigmas)-1:
s=r'%i $\mu$ strain'%(
sigmas[i]*(10**6))
else: s='%i'%(sigmas[i]*(10**6))
axes[i][-1].annotate(
s=s,
verticalalignment='center',
horizontalalignment='center',
rotation=270,
size=14,xy=(1.20,0.5),
xycoords='axes fraction')
fancy_legend(
fig.axes[0],size=11,nscat=1,ncol=1,
bbox_to_anchor=(-0.4,1))
# vm
fig.axes[len(sigmas)*len(nbins)-len(nbins)].\
set_xticks(np.arange(0.,1.01,0.5))
fig.savefig('tab2.pdf')
fig.savefig('tab2.png')
fig.clf()
def tabular_figs01():
"""
Ehkl vs sin2psi plot
y (sigmas)
x (nbins)
"""
fig=plt.figure()
axes=[]
for i in range(len(sigmas)):
axes.append([])
for j in range(len(nbins)):
ax=fig.add_subplot(gs[i,j])
axes[i].append(ax)
ax.locator_params(nbins=4);
if sigmas[i]==0: iscatter=False
else: iscatter= True
rst = main(
## main variables
sigma=sigmas[i], psi_nbin=nbins[j],
## minor
sin2psimx=0.5, iscatter=iscatter,
istep=2,iplot=False,dec_inv_frq=1,
dec_interp=1)
model_rs, sig11, sig22, dsa11,dsa22,\
raw_psis,raw_vfs, raw_sfs,\
full_Ei,dec_intp = rst
# raise IOError
x = sin2psi_opt(model_rs.psis, 1)
ax.plot(
x, model_rs.tdat[0]*1e6,'k.',
label=\
r'$\tilde{ \langle\varepsilon^e \rangle}^G$')
ax.plot(
x, model_rs.Ei[0]*1e6,'kx',
label=\
r'$\mathbb{F}_{ij} \sigma^\mathrm{RS}_{ij}$')
x = sin2psi_opt(raw_psis, 1)
# ax.plot(x, full_Ei[0]*1e6,'k-') ## continuous Ei
## ax.plot(raw_psis, dec_intp[0][0]*1e6,'k--')
## True Ei? = F_ij * <sigma>^c_{ij}
model_rs.psis = raw_psis.copy()
model_rs.npsi = len(model_rs.psis)
model_rs.cffs = raw_sfs.copy()
## plug the weight average stress
model_rs.sigma=[sig11, sig22, 0, 0, 0, 0]
model_rs.calc_Ei()
ax.plot(
x,model_rs.Ei[0]*1e6,'k-',
label=r'$\mathbb{F}_{ij} \langle\sigma\rangle^c_{ij}$'
)
if i==len(sigmas)-1 and j==0:
deco(ax=ax,iopt=0,hkl=None,ipsi_opt=1)
else:
mpl_lib.rm_lab(ax,axis='x')
mpl_lib.rm_lab(ax,axis='y')
for j in range(len(nbins)):
if j==0: s=r'$N^\psi$=%i'%nbins[j]
else:
s = '%i'%nbins[j]
axes[0][j].annotate(
s=s,horizontalalignment='center',
size=14,xy=(0.5,1.2),
xycoords='axes fraction')
for i in range(len(sigmas)):
if i==0:
s=r'$s^\mathrm{CSE}$=%i'%(
sigmas[i]*(10**6))
elif i==len(sigmas)-1:
s=r'%i $\mu$ strain'%(
sigmas[i]*(10**6))
else: s='%i'%(sigmas[i]*(10**6))
axes[i][-1].annotate(
s=s,
verticalalignment='center',
horizontalalignment='center',
rotation=270,
size=14,xy=(1.20,0.5),
xycoords='axes fraction')
for iax in range(len(fig.axes)):
fig.axes[iax].set_ylim(-1500,)
tune_xy_lim(fig.axes)
tune_x_lim(fig.axes,axis='y')
# plt.annotate(s=r'$\varepsilon^\mathrm{CSE}$ Counting Statistics Error',
# size=20,rotation=90,
# verticalalignment='center',
# horizontalalignment='center',
# xy=(1.01,0.5),
# xycoords='figure fraction')
# plt.annotate(s=r'Number of $\psi$',size=20,
# horizontalalignment='center',
# xy=(0.6,0.93),
# xycoords='figure fraction')
fig.axes[len(sigmas)*len(nbins)-len(nbins)].\
set_xticks(np.arange(-0.5,0.501,0.5))
fancy_legend(
fig.axes[0],size=11,nscat=1,ncol=1,
bbox_to_anchor=(-0.4,1))
fig.savefig('tab.pdf')
fig.savefig('tab.png')
fig.clf()
def DEC_evol(steps=[0, 6, 9, 19]):
"""
Draw evolution of DECs with respect to plastic deformation
"""
from MP.mat import mech # mech is a module
FlowCurve = mech.FlowCurve
from mst_ex import use_intp_sfig, return_vf
from rs import ResidualStress, filter_psi3
model_rs = ResidualStress(mod_ext=None,
fnmod_ig='igstrain_fbulk_ph1.out',
fnmod_sf='igstrain_fbulk_ph1.out',
i_ip=1)
fc = FlowCurve(name='Weighted Average')
fc.get_model(fn='STR_STR.OUT')
fc.get_eqv()
phi = model_rs.dat_model.phi
nphi = model_rs.dat_model.nphi
psi = model_rs.dat_model.psi
sin2psi = sin2psi_opt(psi, 1)
DEC_raw = model_rs.dat_model.sf.copy()
vf_raw, dum = return_vf()
DEC_raw, psi, sin2psi, vf_raw = filter_psi3(sin2psi.copy(), [-0.5, 0.5],
DEC_raw, psi.copy(),
sin2psi.copy(), vf_raw.copy())
## filter DEC_raw
DEC_raw[DEC_raw == 0] = np.nan
evm = fc.epsilon_vm.copy()
nstp = fc.nstp
fig = plt.figure()
# steps = [0,3,5,9]
nstp = len(steps)
gs = gridspec.GridSpec(nstp,
nphi,
wspace=0,
hspace=0,
left=0.25,
right=0.8,
top=0.8)
axes = []
axev = []
axs = []
axv = []
for i in xrange(len(steps)):
axes.append([])
axev.append([])
for j in xrange(nphi):
ax = fig.add_subplot(gs[i, j])
av = ax.twinx()
axs.append(ax)
axes[i].append(ax)
axv.append(av)
axev[i].append(av)
ax.locator_params(nbins=4)
lab1 = r'$\mathbb{F}_{11}$'
lab2 = r'$\mathbb{F}_{22}$'
ax.plot(sin2psi,
DEC_raw[steps[i], 0, j, :] * 1e6,
'k-',
label=lab1)
ax.plot(sin2psi,
DEC_raw[steps[i], 1, j, :] * 1e6,
'-',
color='gray',
label=lab2)
av.plot(sin2psi,
vf_raw[steps[i], j, :],
color='gray',
linewidth=3,
alpha=0.7)
ax.plot(0,
0,
'-',
color='gray',
linewidth=3,
alpha=0.7,
label='Vol. F')
if i == len(steps) - 1 and j == 0: pass
else:
mpl_lib.rm_lab(ax, axis='x')
mpl_lib.rm_lab(ax, axis='y')
if i == len(steps) - 1 and j == nphi - 1: pass
else:
mpl_lib.rm_lab(av, axis='x')
mpl_lib.rm_lab(av, axis='y')
tune_xy_lim(axs)
tune_x_lim(axs, axis='y')
tune_xy_lim(axv)
tune_x_lim(axv, axis='y')
for i in xrange(len(steps)): #range(nstp):
if i == 0: s = r'$\bar{E}^{VM}=%.2f$' % fc.epsilon_vm[steps[i]]
else: s = r'$%.2f$' % fc.epsilon_vm[steps[i]]
axes[i][-1].annotate(s=s,
verticalalignment='center',
horizontalalignment='center',
rotation=270,
size=10,
xy=(1.60, 0.5),
xycoords='axes fraction')
for j in xrange(nphi):
s = r'$\phi=%.1f^\circ{}$' % (phi[j] * 180. / np.pi)
axes[0][j].annotate(s=s,
horizontalalignment='center',
size=10,
xy=(0.5, 1.2),
xycoords='axes fraction')
fancy_legend(axes[0][0],
size=10,
nscat=1,
ncol=1,
bbox_to_anchor=(-0.1, 1))
deco(ax=axes[-1][0], ft=10, iopt=1, hkl='211', ipsi_opt=1)
deco(ax=axev[-1][-1], ft=10, iopt=7, hkl='211', ipsi_opt=1)
axev[-1][-1].grid('off')
for i in xrange(len(axv)):
axv[i].set_ylim(0., 0.3)
axv[i].locator_params(nbins=4)
fig.savefig("dec_evol.pdf")
fig.savefig("dec_evol.png")
def tabular_figs02():
fig = plt.figure()
axes = []
for i in xrange(len(sigmas)):
axes.append([])
for j in xrange(len(nbins)):
if sigmas[i] == 0: iscatter = False
else: iscatter = True
rst = main(
## main variables
sigma=sigmas[i],
psi_nbin=nbins[j],
## minor
sin2psimx=0.5,
iscatter=iscatter,
iplot=False,
dec_inv_frq=1,
dec_interp=1)
model_rs, flow_weight, flow_dsa = rst
ax = fig.add_subplot(gs[i, j])
ax.locator_params(nbins=4)
axes[i].append(ax)
# von mises
lab1 = r'$(\langle \sigma^c \rangle)^{VM}$'
lab2 = r'$(\sigma^\mathrm{RS})^{VM}$'
ax.plot(flow_weight.epsilon_vm,
flow_weight.sigma_vm,
'k-',
label=lab1)
ax.plot(flow_dsa.epsilon_vm, flow_dsa.sigma_vm, 'kx', label=lab2)
if i == len(sigmas) - 1 and j == 0:
#VM Deco
axes_label.__vm__(ax=ax, ft=10)
else:
mpl_lib.rm_lab(ax, axis='x')
mpl_lib.rm_lab(ax, axis='y')
for iax in xrange(len(fig.axes)):
fig.axes[iax].set_ylim(0, )
fig.axes[iax].set_xlim(0, )
tune_xy_lim(fig.axes)
tune_x_lim(fig.axes, axis='y')
## annotations
for j in xrange(len(nbins)):
if j == 0: s = r'$N^\psi$=%i' % nbins[j]
else:
s = '%i' % nbins[j]
axes[0][j].annotate(s=s,
horizontalalignment='center',
size=14,
xy=(0.5, 1.2),
xycoords='axes fraction')
for i in xrange(len(sigmas)):
if i == 0:
s = r'$s^\mathrm{CSE}$=%i' % (sigmas[i] * (10**6))
elif i == len(sigmas) - 1:
s = r'%i $\mu$ strain' % (sigmas[i] * (10**6))
else:
s = '%i' % (sigmas[i] * (10**6))
axes[i][-1].annotate(s=s,
verticalalignment='center',
horizontalalignment='center',
rotation=270,
size=14,
xy=(1.20, 0.5),
xycoords='axes fraction')
fancy_legend(fig.axes[0],
size=11,
nscat=1,
ncol=1,
bbox_to_anchor=(-0.4, 1))
# vm
fig.axes[len(sigmas)*len(nbins)-len(nbins)].\
set_xticks(np.arange(0.,1.01,0.5))
fig.savefig('tab2.pdf')
fig.savefig('tab2.png')
fig.clf()
def DEC_intp(ss=[1, 2, 4], intps=[0, 3, 4], inds=[79, 90, 120]):
"""
DEC interpolation comparison
intps=0: NN (piecewise linear interpolation)
=1: Assign nearest data
=2: Cubic
=3: Quadratic
=4: Linear fit
=5: Poly2
=6: poly3
=7: Place-holder for power law fit
=8: zero
=9: slinear
inds=[70,40]
"""
from MP.mat import mech # mech is a module
FlowCurve = mech.FlowCurve
from mst_ex import use_intp_sfig, return_vf
from rs import ResidualStress
model_rs = ResidualStress(mod_ext=None,
fnmod_ig='igstrain_fbulk_ph1.out',
fnmod_sf='igstrain_fbulk_ph1.out',
i_ip=1)
psi = model_rs.dat_model.psi
sin2psi = sin2psi_opt(psi, 1)
DEC_raw = model_rs.dat_model.sf.copy()
fc = FlowCurve(name='Weighted Average')
fc.get_model(fn='STR_STR.OUT')
fc.get_eqv()
evm = fc.epsilon_vm.copy()
fig = plt.figure()
gs = gridspec.GridSpec(len(ss),
len(intps),
wspace=0,
hspace=0,
left=0.25,
right=0.8,
top=0.8)
axes = []
ms = ['o', 'x', '+', 'd', '>', 't']
for i in xrange(len(ss)):
axes.append([])
for j in xrange(len(intps)):
## interpolated DECs
print i, j
print ss[i], intps[j]
# if i==1 and j==0:
# raise IOError
_sf_, _ig_ = use_intp_sfig(ss[i],
iopt=intps[j],
iplot=False,
iwgt=False)
dum = _sf_.sf.swapaxes(-1, -2).swapaxes(-2, -3)
DEC_intp = dum.copy()
ax = fig.add_subplot(gs[i, j])
ax.locator_params(nbins=4)
axes[i].append(ax)
for k in xrange(len(inds)):
ind = inds[k]
val_sin2psi = sin2psi[ind]
val_psi = psi[ind] * 180. / np.pi
lab1, lab2 = None, None
if k == 0:
lab1 = r'Actual $\mathbb{F}_{ij}$'
lab2 = r'Interpolated $\mathbb{F}^{\ I}_{ij}$'
y_raw = DEC_raw[:, 0, 0, ind] * 1e6
ax.plot(evm, y_raw, 'k-', label=lab1)
y_intp = DEC_intp[:, 0, 0, ind] * 1e12
ax.plot(evm, y_intp, 'k--', label=lab2)
ax.fill_between(evm,
y_raw,
y_intp,
facecolor='gray',
alpha=0.5)
ax.plot(evm[::ss[i]],
DEC_raw[:, 0, 0, ind][::ss[i]] * 1e6,
ms[k],
mfc='None',
mec='black',
label=r'$\psi=%2.0f^\circ{}$' % val_psi)
if i == len(ss) - 1 and j == 0:
pass
else:
mpl_lib.rm_lab(ax, axis='x')
mpl_lib.rm_lab(ax, axis='y')
deco(ax=axes[-1][0], ft=10, iopt=6, ij=[1, 1])
axes[-1][0].grid('off')
tune_xy_lim(fig.axes)
tune_x_lim(fig.axes, axis='y')
## annotations
for j in xrange(len(intps)):
if j == 0: s = 'Piecewise'
if j == 1: s = 'Quadratic'
if j == 2: s = 'Linear fit'
if j == 3: s = 'Power law fit'
axes[0][j].annotate(s=s,
horizontalalignment='center',
size=10,
xy=(0.5, 1.2),
xycoords='axes fraction')
for i in xrange(len(ss)):
s = r'$f^{\ \mathbb{F}}=^1/_{%i}$' % ss[i]
axes[i][-1].annotate(s=s,
verticalalignment='center',
horizontalalignment='center',
rotation=270,
size=10,
xy=(1.20, 0.5),
xycoords='axes fraction')
fancy_legend(axes[0][0],
size=10,
nscat=1,
ncol=1,
bbox_to_anchor=(-0.1, 1))
fig.savefig('dec_intp.pdf')
fig.savefig('dec_intp.png')
return model_rs
def tabular_figs01():
"""
Ehkl vs sin2psi plot
y (sigmas)
x (nbins)
"""
fig = plt.figure()
axes = []
for i in xrange(len(sigmas)):
axes.append([])
for j in xrange(len(nbins)):
ax = fig.add_subplot(gs[i, j])
axes[i].append(ax)
ax.locator_params(nbins=4)
if sigmas[i] == 0: iscatter = False
else: iscatter = True
rst = main(
## main variables
sigma=sigmas[i],
psi_nbin=nbins[j],
## minor
sin2psimx=0.5,
iscatter=iscatter,
istep=2,
iplot=False,
dec_inv_frq=1,
dec_interp=1)
model_rs, sig11, sig22, dsa11,dsa22,\
raw_psis,raw_vfs, raw_sfs,\
full_Ei,dec_intp = rst
# raise IOError
x = sin2psi_opt(model_rs.psis, 1)
ax.plot(
x, model_rs.tdat[0]*1e6,'k.',
label=\
r'$\tilde{ \langle\varepsilon^e \rangle}^G$')
ax.plot(
x, model_rs.Ei[0]*1e6,'kx',
label=\
r'$\mathbb{F}_{ij} \sigma^\mathrm{RS}_{ij}$')
x = sin2psi_opt(raw_psis, 1)
# ax.plot(x, full_Ei[0]*1e6,'k-') ## continuous Ei
## ax.plot(raw_psis, dec_intp[0][0]*1e6,'k--')
## True Ei? = F_ij * <sigma>^c_{ij}
model_rs.psis = raw_psis.copy()
model_rs.npsi = len(model_rs.psis)
model_rs.cffs = raw_sfs.copy()
## plug the weight average stress
model_rs.sigma = [sig11, sig22, 0, 0, 0, 0]
model_rs.calc_Ei()
ax.plot(x,
model_rs.Ei[0] * 1e6,
'k-',
label=r'$\mathbb{F}_{ij} \langle\sigma\rangle^c_{ij}$')
if i == len(sigmas) - 1 and j == 0:
deco(ax=ax, iopt=0, hkl=None, ipsi_opt=1)
else:
mpl_lib.rm_lab(ax, axis='x')
mpl_lib.rm_lab(ax, axis='y')
for j in xrange(len(nbins)):
if j == 0: s = r'$N^\psi$=%i' % nbins[j]
else:
s = '%i' % nbins[j]
axes[0][j].annotate(s=s,
horizontalalignment='center',
size=14,
xy=(0.5, 1.2),
xycoords='axes fraction')
for i in xrange(len(sigmas)):
if i == 0:
s = r'$s^\mathrm{CSE}$=%i' % (sigmas[i] * (10**6))
elif i == len(sigmas) - 1:
s = r'%i $\mu$ strain' % (sigmas[i] * (10**6))
else:
s = '%i' % (sigmas[i] * (10**6))
axes[i][-1].annotate(s=s,
verticalalignment='center',
horizontalalignment='center',
rotation=270,
size=14,
xy=(1.20, 0.5),
xycoords='axes fraction')
for iax in xrange(len(fig.axes)):
fig.axes[iax].set_ylim(-1500, )
tune_xy_lim(fig.axes)
tune_x_lim(fig.axes, axis='y')
# plt.annotate(s=r'$\varepsilon^\mathrm{CSE}$ Counting Statistics Error',
# size=20,rotation=90,
# verticalalignment='center',
# horizontalalignment='center',
# xy=(1.01,0.5),
# xycoords='figure fraction')
# plt.annotate(s=r'Number of $\psi$',size=20,
# horizontalalignment='center',
# xy=(0.6,0.93),
# xycoords='figure fraction')
fig.axes[len(sigmas)*len(nbins)-len(nbins)].\
set_xticks(np.arange(-0.5,0.501,0.5))
fancy_legend(fig.axes[0],
size=11,
nscat=1,
ncol=1,
bbox_to_anchor=(-0.4, 1))
fig.savefig('tab.pdf')
fig.savefig('tab.png')
fig.clf()