def corrEtotplot(ax,Eact,Ecmp,a,b,label):
d1 = combinenparrayrange(Eact, a, b, Eidx)
d2 = combinenparrayrange(Ecmp, a, b, Eidx)
rmse = gt.calculaterootmeansqrerror(d1, d2)
slope, intercept, r_value, p_value, std_err = st.linregress(d1, d2)
lin = 'Slope: ' + '%s' % float('%.3g' % slope) + ' Int.: ' + '%s' % float('%.3g' % intercept) + ' $r^2$: ' + '%s' % float('%.3g' % r_value**2)
ax.plot(d1,d1,color='black',label='DFT',linewidth=2,)
ax.scatter(d1, d2, color='red',label=label + ' RMSE ' + '%s' % float('%.3g' % rmse) + ' kcal/mol ' + lin, marker=r"o",s=50)
# Set Limits
ax.set_xlim([d1.min(),d1.max()])
ax.set_ylim([d1.min(),d1.max()])
font = {'family': 'Bitstream Vera Sans',
'weight': 'heavy',
'size': 24}
ax.set_ylabel('$E_{cmp}$',fontdict=font)
ax.set_xlabel('$E_{ref}$',fontdict=font)
ax.legend(bbox_to_anchor=(0.01, 0.99), loc=2, borderaxespad=0., fontsize=10)
t_formatter = matplotlib.ticker.ScalarFormatter(useOffset=False)
ax.xaxis.set_major_formatter(t_formatter)
ax.yaxis.set_major_formatter(t_formatter)
def graphEdiffDelta2D(ax, data1, data2, Na):
#data1 = gt.convert * data1
#data2 = gt.convert * data2
x, y, z, d = gt.calculateelementdiff2D(data1)
x2, y2, z2, d2 = gt.calculateelementdiff2D(data2)
RMSE = gt.calculaterootmeansqrerror(d, d2) / float(Na)
print('dataz1:', z)
print('dataz2:', z2)
z = np.abs(z - z2)
print('zdiff:', z)
# Set up a regular grid of interpolation points
xi, yi = np.linspace(x.min(), x.max(),
100), np.linspace(y.min(), y.max(), 100)
xi, yi = np.meshgrid(xi, yi)
# Interpolate
rbf = scipy.interpolate.Rbf(x, y, z, function='linear')
zi = rbf(xi, yi)
im = ax.imshow(zi,
vmin=z.min(),
vmax=z.max(),
origin='lower',
extent=[x.min(), x.max(),
y.min(), y.max()])
ax.set_title('Difference Plot (symmetric)\nRMSE: ' +
"{:.5f}".format(RMSE) + 'kcal/mol/atom Atoms: ' + str(Na),
fontsize=14)
# plt.xlabel('Structure')
# plt.ylabel('Structure')
ax.scatter(x, y, c=z, vmin=z.min(), vmax=z.max())
ax.set_xlim([-0.02 * x.max(), x.max() + 0.02 * x.max()])
ax.set_ylim([-0.02 * y.max(), y.max() + 0.02 * y.max()])
#ax.plot([x.min(),x.max()],[y.min(),x.max()],color='red',linewidth=4)
ax.grid(True)
return im
def graphEdiffDelta2D(ax, data1, data2, Na):
#data1 = gt.convert * data1
#data2 = gt.convert * data2
x, y, z, d = gt.calculateelementdiff2D(data1)
x2, y2, z2, d2 = gt.calculateelementdiff2D(data2)
RMSE = gt.calculaterootmeansqrerror(d, d2) / float(Na)
print('Number of atoms: ' + str(Na))
print('RMSE: ' + str(RMSE) + ' kcal/mol/atom')
print('RMSE: ' + str(float(Na) * RMSE) + ' kcal/mol')
z = np.abs(z - z2)
C = data1.shape[0]
mat = np.ndarray(shape=(C, C), dtype=float)
for i in x:
for j in y:
I = int(i)
J = int(j)
mat[J, I] = z[J + I * C]
#get discrete colormap
cmap = plt.get_cmap('RdBu', np.max(mat) - np.min(mat) + 1)
# Show mat
im = ax.matshow(mat, vmin=np.min(mat), vmax=np.max(mat))
cmap = plt.cm.jet
norm = plt.Normalize(mat.min(), mat.max())
rgba = cmap(norm(mat))
rgba[range(C), range(C), :3] = 1, 1, 1
ax.imshow(rgba, interpolation='nearest')
# Plot center line
#ax.plot([x.min()-0.5,x.max()+0.5],[y.min()-0.5,x.max()+0.5],color='black',linewidth=4,alpha=0.9)
# Set Limits
ax.set_xlim([-0.02 * x.max(), x.max() + 0.02 * x.max()])
ax.set_ylim([-0.02 * y.max(), y.max() + 0.02 * y.max()])
ax.xaxis.tick_bottom()
return im
def graphEdiffDelta2D(ax, data1, data2, Na):
#data1 = gt.convert * data1
#data2 = gt.convert * data2
x, y, z, d = gt.calculateelementdiff2D(data1)
x2, y2, z2, d2 = gt.calculateelementdiff2D(data2)
RMSE = gt.calculaterootmeansqrerror(d,d2) / float(Na)
print ('Number of atoms: ' + str(Na))
print ('RMSE: ' + str(RMSE) + ' kcal/mol/atom')
print ('RMSE: ' + str(float(Na)*RMSE) + ' kcal/mol')
z = np.abs(z - z2)
C = data1.shape[0]
mat = np.ndarray(shape=(C,C),dtype=float)
for i in x:
for j in y:
I = int(i)
J = int(j)
mat[J,I] = z[J+I*C]
#get discrete colormap
cmap = plt.get_cmap('RdBu', np.max(mat)-np.min(mat)+1)
# Show mat
im = ax.matshow(mat,vmin = np.min(mat), vmax = np.max(mat))
cmap = plt.cm.jet
norm = plt.Normalize(mat.min(), mat.max())
rgba = cmap(norm(mat))
rgba[range(C), range(C), :3] = 1, 1, 1
ax.imshow(rgba, interpolation='nearest')
# Plot center line
#ax.plot([x.min()-0.5,x.max()+0.5],[y.min()-0.5,x.max()+0.5],color='black',linewidth=4,alpha=0.9)
# Set Limits
ax.set_xlim([-0.02*x.max(),x.max()+0.02*x.max()])
ax.set_ylim([-0.02*y.max(),y.max()+0.02*y.max()])
ax.xaxis.tick_bottom()
return im
def Ecorrplot (ax1,Eact,Ecmp,mlbl,color,lab=False):
mx = Eact.max()
mn = Eact.min()
if lab:
ax1.plot((mn, mx), (mn, mx), color='black',label='DFT', linewidth=5)
else:
ax1.plot((mn, mx), (mn, mx), color='black', linewidth=5)
rmse = gt.calculaterootmeansqrerror(Eact, Ecmp)
ax1.scatter(Eact, Ecmp, marker=r'o', color=color, label= mlbl + ' RMSE: ' + "{:.3f}".format(rmse) + ' kcal/mol', linewidth=1)
ax1.set_xlim([mn,mx])
ax1.set_ylim([mn,mx])
#ax1.set_title("title)
ax1.set_ylabel('$\Delta E_{cmp}$ (kcal/mol)')
ax1.set_xlabel('$\Delta E_{ref}$ (kcal/mol)')
ax1.legend(bbox_to_anchor=(0.01, 0.99), loc=2, borderaxespad=0., fontsize=16)
def graphEdiffDelta2D(ax, data1, data2, Na):
data1 = gt.convert * data1
data2 = gt.convert * data2
x, y, z, d = gt.calculateelementdiff2D(data1)
x2, y2, z2, d2 = gt.calculateelementdiff2D(data2)
RMSE = gt.calculaterootmeansqrerror(d, d2) / float(Na)
print("dataz1:", z)
print("dataz2:", z2)
z = np.abs(z - z2)
print("zdiff:", z)
# Set up a regular grid of interpolation points
xi, yi = np.linspace(x.min(), x.max(), 100), np.linspace(y.min(), y.max(), 100)
xi, yi = np.meshgrid(xi, yi)
# Interpolate
rbf = scipy.interpolate.Rbf(x, y, z, function="linear")
zi = rbf(xi, yi)
im = ax.imshow(zi, vmin=z.min(), vmax=z.max(), origin="lower", extent=[x.min(), x.max(), y.min(), y.max()])
ax.set_title(
"Difference Plot (symmetric)\nRMSE: " + "{:.5f}".format(RMSE) + "kcal/mol/atom Atoms: " + str(natm), fontsize=14
)
# plt.xlabel('Structure')
# plt.ylabel('Structure')
ax.scatter(x, y, c=z, vmin=z.min(), vmax=z.max())
ax.set_xlim([-0.02 * x.max(), x.max() + 0.02 * x.max()])
ax.set_ylim([-0.02 * y.max(), y.max() + 0.02 * y.max()])
# ax.plot([x.min(),x.max()],[y.min(),x.max()],color='red',linewidth=4)
ax.grid(True)
return im
def corrEtotplot(ax, Eact, Ecmp, a, b, label):
d1 = combinenparrayrange(Eact, a, b, Eidx)
d2 = combinenparrayrange(Ecmp, a, b, Eidx)
rmse = gt.calculaterootmeansqrerror(d1, d2)
slope, intercept, r_value, p_value, std_err = st.linregress(d1, d2)
lin = (
"Slope: "
+ "%s" % float("%.3g" % slope)
+ " Int.: "
+ "%s" % float("%.3g" % intercept)
+ " $r^2$: "
+ "%s" % float("%.3g" % r_value ** 2)
)
ax.plot(d1, d1, color="black", label="DFT", linewidth=2)
ax.scatter(
d1,
d2,
color="red",
label=label + " RMSE " + "%s" % float("%.3g" % rmse) + " kcal/mol " + lin,
marker=r"o",
s=50,
)
# Set Limits
ax.set_xlim([d1.min(), d1.max()])
ax.set_ylim([d1.min(), d1.max()])
font = {"family": "Bitstream Vera Sans", "weight": "heavy", "size": 24}
ax.set_ylabel("$E_{cmp}$", fontdict=font)
ax.set_xlabel("$E_{ref}$", fontdict=font)
ax.legend(bbox_to_anchor=(0.01, 0.99), loc=2, borderaxespad=0.0, fontsize=10)
t_formatter = matplotlib.ticker.ScalarFormatter(useOffset=False)
ax.xaxis.set_major_formatter(t_formatter)
ax.yaxis.set_major_formatter(t_formatter)
# Print some data from the NeuroChem
print( '1) Number of Atoms Loaded: ' + str(nc.getNumAtoms()) )
print( '1) Number of Confs Loaded: ' + str(nc.getNumConfs()) )
# Compute Energies of Conformations
print('Computing energies...')
_t1b = tm.time()
Ecmp_t = nc.computeEnergies()
_t2b = (tm.time() - _t1b) * 1000.0
print('Computation complete. Time: ' + "{:.4f}".format(_t2b) + 'ms')
Ecmp_t = setmaxE(Eact_t, Ecmp_t, 300.0)
Eact_t = setmaxE(Eact_t, Eact_t, 300.0)
tNa = nc.getNumAtoms()
err.append(gt.hatokcal * gt.calculaterootmeansqrerror(np.array(Eact_t, dtype=float),np.array(Ecmp_t, dtype=float)) / float(tNa))
sze.append(float(len(Eact_t)))
time += _t2b
Eidx.append(N)
Emin.append(gt.hatokcal * np.array( Ecmp_t ).min())
Efle.append(i)
N += 1
Ecmp.append( gt.hatokcal * np.array( Ecmp_t ) )
Eact.append( gt.hatokcal * np.array( Eact_t ) )
err = np.array(err, dtype=float)
sze = np.array(sze, dtype=float)
# n = 0
# m = 9
# Ecmp1 = gt.hatokcal * Ecmp1[n:m]
# Eact = gt.hatokcal * Eact[n:m]
Ecmp1 = gt.hatokcal * Ecmp1
Eact = gt.hatokcal * Eact
# Eact2 = gt.hatokcal * Eact2
IDX = np.arange(0, Eact.shape[0], 1, dtype=float) + 1
# Ecmp1 = Ecmp1 - Ecmp1.min()
# Eact = Eact - Eact.min()
# Eact2 = Eact2 - Eact2.min()
rmse1 = gt.calculaterootmeansqrerror(Eact, Ecmp1)
# rmse2 = gt.calculaterootmeansqrerror(Eact,Eact2)
print("Spearman corr. 1: " + "{:.3f}".format(st.spearmanr(Ecmp1, Eact)[0]))
# print ( "Spearman corr. 2: " + "{:.3f}".format(st.spearmanr(Eact2,Eact)[0]) )
# plt.plot (IDX, Eact, '-', marker=r'o', color='black',label='DFT', linewidth=2, markersize=10)
# plt.plot (IDX, Ecmp1, ':', marker=r'D', color='red', label='ANI-1 RMSE: ' + "{:.7f}".format(rmse1) + ' kcal/mol', linewidth=4, markersize=8)
plt.plot(IDX, Eact, "-", color="black", label="DFT", linewidth=3)
plt.plot(IDX, Ecmp1, "--", color="red", label="ANI-1 RMSE: " + "{:.3f}".format(rmse1) + " kcal/mol", linewidth=3)
# plt.plot (IDX, Eact2, '--', color='blue', label='DFTB RMSE: ' + "{:.3f}".format(rmse2) + ' kcal/mol', linewidth=2)
plt.title("BO MD Trajectory - Ranolazine")
plt.ylabel("$\Delta$E calculated (kcal/mol)")
plt.xlabel("Step number")
#n = 0
#m = 9
#Ecmp1 = gt.hatokcal * Ecmp1[n:m]
#Eact = gt.hatokcal * Eact[n:m]
Ecmp1 = gt.hatokcal * Ecmp1
Eact = gt.hatokcal * Eact
#Eact2 = gt.hatokcal * Eact2
IDX = np.arange(0,Eact.shape[0],1,dtype=float) + 1
#Ecmp1 = Ecmp1 - Ecmp1.min()
#Eact = Eact - Eact.min()
#Eact2 = Eact2 - Eact2.min()
rmse1 = gt.calculaterootmeansqrerror(Eact,Ecmp1)
#rmse2 = gt.calculaterootmeansqrerror(Eact,Eact2)
print ( "Spearman corr. 1: " + "{:.3f}".format(st.spearmanr(Ecmp1,Eact)[0]) )
#print ( "Spearman corr. 2: " + "{:.3f}".format(st.spearmanr(Eact2,Eact)[0]) )
#plt.plot (IDX, Eact, '-', marker=r'o', color='black',label='DFT', linewidth=2, markersize=10)
#plt.plot (IDX, Ecmp1, ':', marker=r'D', color='red', label='ANI-1 RMSE: ' + "{:.7f}".format(rmse1) + ' kcal/mol', linewidth=4, markersize=8)
plt.plot (IDX, Eact, '-', color='black',label='DFT', linewidth=3)
plt.plot (IDX, Ecmp1, '--', color='red', label='ANI-1 RMSE: ' + "{:.3f}".format(rmse1) + ' kcal/mol', linewidth=3)
#plt.plot (IDX, Eact2, '--', color='blue', label='DFTB RMSE: ' + "{:.3f}".format(rmse2) + ' kcal/mol', linewidth=2)
plt.title("BO MD Trajectory - Ranolazine")
plt.ylabel('$\Delta$E calculated (kcal/mol)')
plt.xlabel('Step number')
dir = '/Research/ANN-Test-Data/GDB-11-W98XD-6-31gd/train_05/'
dir2 = '/Research/ANN-Test-Data/GDB-11-W98XD-6-31gd/train_06/'
#file = 'RMSEperATOM.dat'
file = 'gdb11_s05-99_test.dat_graph'
data1 = gt.getfltsfromfile('/home/' + user + dir + file, [0])
data2 = gt.getfltsfromfile('/home/' + user + dir + file, [1])
data3 = gt.getfltsfromfile('/home/' + user + dir + file, [2])
data4 = gt.getfltsfromfile('/home/' + user + dir2 + file, [2])
data2 = gt.calculateelementdiff(data2)
data3 = gt.calculateelementdiff(data3)
data4 = gt.calculateelementdiff(data4)
rmse5 = gt.calculaterootmeansqrerror(data2[:,1],data3[:,1]) / 63.0
rmse6 = gt.calculaterootmeansqrerror(data2[:,1],data4[:,1]) / 63.0
print('Datasize: ' + str(data1.shape[0]))
font = {'family' : 'Bitstream Vera Sans',
'weight' : 'normal',
'size' : 14}
plt.rc('font', **font)
#print(data2)
plt.plot(data2[:,0], data2[:,1], color='black', label='wB97X/6-31G*',linewidth=2)
plt.scatter(data2[:,0], data2[:,1], color='black',linewidth=4)
plt.plot(data3[:,0], data3[:,1],'r--', color='blue', label='ANN - GDB-5 RMSE: ' + "{:.6f}".format(rmse5) + "eV/atom",linewidth=2)
print("1) Number of Confs Loaded: " + str(nc.getNumConfs()))
# Compute Energies of Conformations
print("Computing energies...")
_t1b = tm.time()
Ecmp_t = nc.computeEnergies()
_t2b = (tm.time() - _t1b) * 1000.0
print("Computation complete. Time: " + "{:.4f}".format(_t2b) + "ms")
Ecmp_t = setmaxE(Eact_t, Ecmp_t, 300.0)
Eact_t = setmaxE(Eact_t, Eact_t, 300.0)
tNa = nc.getNumAtoms()
err.append(
gt.hatokcal
* gt.calculaterootmeansqrerror(np.array(Eact_t, dtype=float), np.array(Ecmp_t, dtype=float))
/ float(tNa)
)
sze.append(float(len(Eact_t)))
time += _t2b
Eidx.append(N)
Emin.append(gt.hatokcal * np.array(Ecmp_t).min())
Efle.append(i)
N += 1
Ecmp.append(gt.hatokcal * np.array(Ecmp_t))
Eact.append(gt.hatokcal * np.array(Eact_t))
err = np.array(err, dtype=float)
[1])
data3 = gt.convert * gt.getfltsfromfile('/home/' + user + dir1 + file, ' ',
[2])
#data4 = gt.convert * gt.getfltsfromfile('/home/' + user + dir1 + file2,' ', [1])
#data4 = gt.getfltsfromfile('/home/' + user + dir2 + file,' ', [1])
#data5 = gt.getfltsfromfile('/home/' + user + dir2 + file,' ', [2])
#data5 = gt.getfltsfromfile('/home/' + user + dir3 + file, [2])
#mean = np.mean(data3)
#data2 = data2 - np.mean(data2)
#data3 = data3 - np.mean(data3)
#data4 = data4 - np.mean(data4)
rmse1 = gt.calculaterootmeansqrerror(data2, data3)
#rmse2 = gt.calculaterootmeansqrerror(data2,data4)
#rmse3 = gt.calculaterootmeansqrerror(data2,data5)
print('Datasize: ' + str(data1.shape[0]))
font = {'family': 'Bitstream Vera Sans', 'weight': 'normal', 'size': 14}
plt.rc('font', **font)
data1 = data1
plt.plot(data1, data2, color='black', label='wB97X/6-31G*', linewidth=4)
plt.scatter(data1, data2, color='black', linewidth=3)
plt.scatter(data1, data3, color='blue', linewidth=3)
plt.plot(data1,
def produce_scan(ax,title,xlabel,cnstfile,saefile,nnfdir,dtdir,dt1,dt2,dt3,smin,smax,iscale,ishift):
xyz, typ, Eact, chk = gt.readncdat(dtdir + dt1,np.float32)
xyz2, typ2, Eact2, chk = gt.readncdat(dtdir + dt2)
xyz3, typ3, Eact3, chk = gt.readncdat(dtdir + dt3)
#gt.writexyzfile("/home/jujuman/Dropbox/ChemSciencePaper.AER/TestCases/Dihedrals/4-Cyclohexyl-1-butanol/optimization/dihedral_"+dt1+".xyz",xyz,typ)
#Eact = np.array(Eact)
#Eact2 = np.array(Eact2)
#Eact3 = np.array(Eact3)
# Construct pyNeuroChem classes
nc1 = pync.pyNeuroChem(cnstfile, saefile, nnfdir, 1)
# Set the conformers in NeuroChem
nc1.setConformers(confs=xyz, types=typ)
# Print some data from the NeuroChem
print('1) Number of Atoms Loaded: ' + str(nc1.getNumAtoms()))
print('1) Number of Confs Loaded: ' + str(nc1.getNumConfs()))
# Compute Forces of Conformations
print('Computing energies 1...')
_t1b = tm.time()
Ecmp1 = nc1.energy()
print('Computation complete 1. Time: ' + "{:.4f}".format((tm.time() - _t1b) * 1000.0) + 'ms')
n = smin
m = smax
Ecmp1 = gt.hatokcal * Ecmp1
Eact = gt.hatokcal * Eact
Eact2 = gt.hatokcal * Eact2
Eact3 = gt.hatokcal * Eact3
IDX = np.arange(0, Eact.shape[0], 1, dtype=float) * iscale + ishift
IDX = IDX[n:m]
Eact = Eact[n:m]
Eact2 = Eact2[n:m]
Eact3 = Eact3[n:m]
Ecmp1 = Ecmp1[n:m]
Ecmp1 = Ecmp1 - Ecmp1.min()
Eact = Eact - Eact.min()
Eact2 = Eact2 - Eact2.min()
Eact3 = Eact3 - Eact3.min()
rmse1 = gt.calculaterootmeansqrerror(Eact, Ecmp1)
rmse3 = gt.calculaterootmeansqrerror(Eact, Eact2)
rmse4 = gt.calculaterootmeansqrerror(Eact, Eact3)
print("Spearman corr. 1: " + "{:.3f}".format(st.spearmanr(Ecmp1, Eact)[0]))
print("Spearman corr. 2: " + "{:.3f}".format(st.spearmanr(Eact2, Eact)[0]))
print("Spearman corr. 3: " + "{:.3f}".format(st.spearmanr(Eact3, Eact)[0]))
ax.plot(IDX, Eact, '-', marker=r'o', color='black', label='DFT',
linewidth=2, markersize=7)
ax.plot(IDX, Ecmp1, ':', marker=r'D', color='red', label='ANI-1 RMSE: ' + '%s' % float('%.3g' % rmse1) + ' kcal/mol',
linewidth=2, markersize=5)
ax.plot(IDX, Eact2, ':', marker=r'v', color='blue', label='DFTB RMSE: ' + '%s' % float('%.3g' % rmse3) + ' kcal/mol',
linewidth=2, markersize=5)
ax.plot(IDX, Eact3, ':', marker=r'*', color='orange', label='PM6 RMSE: ' + '%s' % float('%.3g' % rmse4) + ' kcal/mol',
linewidth=2, markersize=7)
#ax.plot(IDX, Eact, color='black', label='DFT', linewidth=3)
#ax.scatter(IDX, Eact, marker='o', color='black', linewidth=4)
th = ax.set_title(title,fontsize=16)
th.set_position([0.5,1.005])
# Set Limits
ax.set_xlim([ IDX.min(),IDX.max()])
ax.set_ylim([Eact.min()-1.0,Eact.max()+1.0])
ax.set_ylabel('$\Delta$E calculated (kcal/mol)')
ax.set_xlabel(xlabel)
ax.legend(bbox_to_anchor=(0.2, 0.98), loc=2, borderaxespad=0., fontsize=14)
def produce_scan(title, xlabel, cnstfile, saefile, nnfdir, dtdir, dt1, smin,
smax, iscale, ishift, atm):
xyz, frc, typ, Eact, chk = gt.readncdatwforce(dtdir + dt1)
xyz = np.asarray(xyz, dtype=np.float32)
xyz = xyz.reshape((xyz.shape[0], len(typ), 3))
print(xyz)
frc = np.asarray(frc, dtype=np.float32)
frc = frc.reshape((frc.shape[0], len(typ), 3))
Eact = np.array(Eact)
# Construct pyNeuroChem classes
nc1 = pync.pyNeuroChem(cnstfile, saefile, nnfdir, 0)
# Set the conformers in NeuroChem
nc1.setConformers(confs=xyz, types=typ)
# Print some data from the NeuroChem
print('1) Number of Atoms Loaded: ' + str(nc1.getNumAtoms()))
print('1) Number of Confs Loaded: ' + str(nc1.getNumConfs()))
# Compute Energies of Conformations
print('Computing energies...')
_t1b = tm.time()
Ecmp1 = nc1.energy()
print('Energy computation complete. Time: ' +
"{:.4f}".format((tm.time() - _t1b) * 1000.0) + 'ms')
# Compute Forces of Conformations
print('Compute forces...')
_t1b = tm.time()
F = nc1.force()
print('Force computation complete. Time: ' +
"{:.4f}".format((tm.time() - _t1b) * 1000.0) + 'ms')
#Fn = np.array(nc1.computeNumericalForces(dr=0.0001))
n = smin
m = smax
Ecmp1 = gt.hatokcal * Ecmp1
Eact = gt.hatokcal * Eact
IDX = np.arange(0, Eact.shape[0], 1, dtype=float) * iscale + ishift
IDX = IDX
Eact = Eact
Ecmp1 = Ecmp1
Ecmp1 = Ecmp1 - Ecmp1.min()
Eact = Eact - Eact.min()
rmse1 = gt.calculaterootmeansqrerror(Eact, Ecmp1)
print("Spearman corr. 1: " + "{:.3f}".format(st.spearmanr(Ecmp1, Eact)[0]))
fig, axes = plt.subplots(nrows=2, ncols=2)
axes.flat[0].plot(IDX, Eact, '-', color='black', label='DFT', linewidth=6)
axes.flat[0].plot(IDX,
Ecmp1,
'--',
color='red',
label='ANI-1',
linewidth=6)
#ax.plot(IDX, Eact, color='black', label='DFT', linewidth=3)
#ax.scatter(IDX, Eact, marker='o', color='black', linewidth=4)
th = axes.flat[0].set_title(title, fontsize=13)
th.set_position([0.5, 1.00])
# Set Limits
axes.flat[0].set_xlim([IDX.min(), IDX.max()])
#axes.flat[0].set_ylim([Eact.min()-1.0,Eact.max()+1.0])
axes.flat[0].set_ylabel('$\Delta$E calculated (kcal/mol)')
axes.flat[0].set_xlabel(xlabel)
axes.flat[0].legend(bbox_to_anchor=(0.2, 0.98),
loc=2,
borderaxespad=0.,
fontsize=12)
#print (Fn)
for i in range(0, 3):
Fq = F[:, :, i][:, atm]
#Fnq = Fn[:,i]
Faq = (1.8897259885789 * np.array(frc)[:, :, i][:, atm])
print(Fq)
print(Faq)
th = axes.flat[i + 1].set_title("Force for atom " + str(atm) +
": coordinate " + str(i),
fontsize=13)
th.set_position([0.5, 1.00])
axes.flat[i + 1].set_xlim([IDX.min(), IDX.max()])
axes.flat[i + 1].plot(IDX,
Faq,
'-',
color='black',
label='DFT',
linewidth=6)
#axes.flat[i+1].plot(IDX, Fnq, '-', color='blue', label='ANI Numerical',
# linewidth=6)
axes.flat[i + 1].plot(IDX,
Fq,
'--',
color='red',
label='ANI Analytical',
linewidth=6)
axes.flat[i + 1].set_ylabel('Force (Ha/A)')
axes.flat[i + 1].set_xlabel(xlabel)
axes.flat[i + 1].legend(bbox_to_anchor=(0.2, 0.98),
loc=2,
borderaxespad=0.,
fontsize=12)
font = {'family': 'Bitstream Vera Sans', 'weight': 'normal', 'size': 12}
plt.rc('font', **font)
plt.show()
mse_t = sqd / float(dpts)
print('|-----------Prediction Complete-----------|')
print(' Epoch error MSE(Ha) -- ')
print(' Test: ' + "{:.5f}".format(mse_t))
print(' Epoch error RMSE(kcal/mol) -- ')
print(' Test: ' + "{:.5f}".format(gt.hatokcal * np.sqrt(mse_t)))
print('|-----------------------------------------|')
Ecmp = gt.hatokcal * np.array(Ecmp, dtype=float)
Eact = gt.hatokcal * np.array(Eact, dtype=float)
# Compute RMSE in kcal/mol
rmse = gt.calculaterootmeansqrerror(Ecmp, Eact)
# End timer
_t1e = tm.time()
print('Computation complete. Time: ' + "{:.4f}".format((_t1e - _t1b)) + 's')
# Output model information
print('RMSE: ' + str(rmse))
# Plot
plt.scatter(Eact,
Ecmp,
label='Baseline RMSE: ' + '%s' % float('%.3g' % rmse) +
' kcal/mol',
color='red')
plt.scatter(Eact, Eact, label='DFT', color='blue')
file = 'pp_02_test.dat_graph'
#file = 'polypep_test.dat_graph'
#file = 'aminoacid_00-12_test.dat_graph'
#file = 'benzamide_conformers-0_test.dat_graph'
#file = 'pentadecane_test.dat_graph'
#file = 'retinolconformer_test.dat_graph'
#data1 = gt.getfltsfromfile('/home/' + user + dir1 + file, ' ', [0])
data0 = gt.getfltsfromfile('/home/' + user + dir1 + file, ' ', [1])
data1 = gt.getfltsfromfile('/home/' + user + dir1 + file, ' ', [2])
data0 = gt.calculateelementdiff(data0)
data1 = gt.calculateelementdiff(data1)
rmse1 = 27.2113825435 * gt.calculaterootmeansqrerror(data0[:,1],data1[:,1]) / 24.0
print('Datasize: ' + str(data1.shape[0]))
font = {'family' : 'Bitstream Vera Sans',
'weight' : 'normal',
'size' : 8}
plt.rc('font', **font)
data = data0
plt.plot(data[:,0], data[:,1], color='black', label='wB97X/6-31G*',linewidth=2)
plt.scatter(data[:,0], data[:,1], color='black',linewidth=4)
data = data1
plt.plot(data[:,0], data[:,1],'r--', color='blue', label='ANN - GDB-7 (6.1$\AA$) RMSE: ' + "{:.6f}".format(rmse1) + "eV/atom",linewidth=2)
file = 'pp_02_test.dat_graph'
#file = 'polypep_test.dat_graph'
#file = 'aminoacid_00-12_test.dat_graph'
#file = 'benzamide_conformers-0_test.dat_graph'
#file = 'pentadecane_test.dat_graph'
#file = 'retinolconformer_test.dat_graph'
#data1 = gt.getfltsfromfile('/home/' + user + dir1 + file, ' ', [0])
data0 = gt.getfltsfromfile('/home/' + user + dir1 + file, ' ', [1])
data1 = gt.getfltsfromfile('/home/' + user + dir1 + file, ' ', [2])
data0 = gt.calculateelementdiff(data0)
data1 = gt.calculateelementdiff(data1)
rmse1 = 27.2113825435 * gt.calculaterootmeansqrerror(data0[:, 1],
data1[:, 1]) / 24.0
print('Datasize: ' + str(data1.shape[0]))
font = {'family': 'Bitstream Vera Sans', 'weight': 'normal', 'size': 8}
plt.rc('font', **font)
data = data0
plt.plot(data[:, 0],
data[:, 1],
color='black',
label='wB97X/6-31G*',
linewidth=2)
plt.scatter(data[:, 0], data[:, 1], color='black', linewidth=4)
Ecmp3 = gt.hatokcal * Ecmp3
Ecmp4 = gt.hatokcal * Ecmp4
Ecmp5 = gt.hatokcal * Ecmp5
Eact = gt.hatokcal * Eact
Eotr = gt.hatokcal * Eotr
Emax = 400.0
Ecmp1 = setmaxE(Eact,Ecmp1,Emax)
Ecmp2 = setmaxE(Eact,Ecmp2,Emax)
Ecmp3 = setmaxE(Eact,Ecmp3,Emax)
Ecmp4 = setmaxE(Eact,Ecmp4,Emax)
Ecmp5 = setmaxE(Eact,Ecmp5,Emax)
Eotr = setmaxE(Eact,Eotr, Emax)
Eact = setmaxE(Eact,Eact, Emax)
rmse1 = gt.calculaterootmeansqrerror(Eact,Eotr)
rmse2 = gt.calculaterootmeansqrerror(Eact,Ecmp1)
rmse3 = gt.calculaterootmeansqrerror(Eact,Ecmp2)
rmse4 = gt.calculaterootmeansqrerror(Eact,Ecmp3)
rmse5 = gt.calculaterootmeansqrerror(Eact,Ecmp4)
rmse6 = gt.calculaterootmeansqrerror(Eact,Ecmp5)
mx = Eact.max()
mn = Eact.min()
#plt.scatter(IDX, Eact, marker='o' , color='black', linewidth=3)
print ( "Spearman corr. DFTB: " + "{:.3f}".format(st.spearmanr(Eotr,Eact)[0]) )
print ( "Spearman corr. TGM 08: " + "{:.3f}".format(st.spearmanr(Ecmp1,Eact)[0]) )
#print ( "Spearman corr. TGM 07: " + "{:.3f}".format(st.spearmanr(Ecmp2,Eact)[0]) )
#print ( "Spearman corr. TGM 06: " + "{:.3f}".format(st.spearmanr(Ecmp3,Eact)[0]) )
data1 = gt.getfltsfromfile('/home/' + user + dir1 + file,' ', [0])
data2 = gt.convert * gt.getfltsfromfile('/home/' + user + dir1 + file,' ', [1])
data3 = gt.convert * gt.getfltsfromfile('/home/' + user + dir1 + file,' ', [2])
#data4 = gt.convert * gt.getfltsfromfile('/home/' + user + dir1 + file2,' ', [1])
#data4 = gt.getfltsfromfile('/home/' + user + dir2 + file,' ', [1])
#data5 = gt.getfltsfromfile('/home/' + user + dir2 + file,' ', [2])
#data5 = gt.getfltsfromfile('/home/' + user + dir3 + file, [2])
#mean = np.mean(data3)
#data2 = data2 - np.mean(data2)
#data3 = data3 - np.mean(data3)
#data4 = data4 - np.mean(data4)
rmse1 = gt.calculaterootmeansqrerror(data2,data3)
#rmse2 = gt.calculaterootmeansqrerror(data2,data4)
#rmse3 = gt.calculaterootmeansqrerror(data2,data5)
print('Datasize: ' + str(data1.shape[0]))
font = {'family' : 'Bitstream Vera Sans',
'weight' : 'normal',
'size' : 14}
plt.rc('font', **font)
data1 = data1
plt.plot(data1, data2, color='black', label='wB97X/6-31G*',linewidth=4)
plt.scatter(data1, data2, color='black',linewidth=3)
"""
# Fit model
clf.fit(X_train, y_train)
# Compute and print r^2 score
print(clf.score(X_test, y_test))
# Store predicted energies
Ecmp = clf.predict(X_test)
Ecmp = gt.hatokcal * (Ecmp)
Eact = gt.hatokcal * (y_test)
# Compute RMSE in kcal/mol
rmse = gt.calculaterootmeansqrerror(Ecmp, Eact)
# End timer
_t1e = tm.time()
print("Computation complete. Time: " + "{:.4f}".format((_t1e - _t1b)) + "s")
# Output model information
print("RMSE: " + str(rmse))
# print(clf.coef_)
# print(clf.intercept_)
# Plot
plt.scatter(Eact, Ecmp, label="Baseline RMSE: " + "%s" % float("%.3g" % rmse) + " kcal/mol", color="red")
plt.scatter(Eact, Eact, label="DFT", color="blue")
plt.title("GDB-2 - CM/MLP energy correlation to DFT")
deltas = gt.hatokcal * np.abs(Ecmp_t - np.array(Eact_t, dtype=float))
Me = max (deltas)
if Me > Herror:
Herror = Me
Wfile = i
Le = min (deltas)
if Le < Lerror:
Lerror = Le
Bfile = i
#print (gt.hatokcal * gt.calculaterootmeansqrerror(np.array(Eact_t, dtype=float),Ecmp_t))
tNa = nc.getNumAtoms()
err.append(gt.hatokcal * gt.calculaterootmeansqrerror(np.array(Eact_t, dtype=float),Ecmp_t) / float(tNa))
sze.append(float(len(Eact_t)))
time += _t2b
Ecmp += Ecmp_t
Eact += Eact_t
#plt_by_index(np.array(Eerr),-1)
Ndps = len(Ecmp)
print ('\nMax Delta (kcal/mol): ' + str(Herror) + ' FILE: ' + Wfile)
print ('Min Delta (kcal/mol): ' + str(Lerror) + ' FILE: ' + Bfile)
print('\nMAXE')
print(ld)
def produce_scan(title, xlabel, cnstfile, saefile, nnfdir, dtdir, dt1, smin, smax, iscale, ishift, atm):
xyz, frc, typ, Eact, chk = gt.readncdatwforce(dtdir + dt1)
xyz = np.asarray(xyz, dtype=np.float32)
xyz = xyz.reshape((xyz.shape[0], len(typ), 3))
print(xyz)
frc = np.asarray(frc, dtype=np.float32)
frc = frc.reshape((frc.shape[0], len(typ), 3))
Eact = np.array(Eact)
# Construct pyNeuroChem classes
nc1 = pync.pyNeuroChem(cnstfile, saefile, nnfdir, 0)
# Set the conformers in NeuroChem
nc1.setConformers(confs=xyz, types=typ)
# Print some data from the NeuroChem
print("1) Number of Atoms Loaded: " + str(nc1.getNumAtoms()))
print("1) Number of Confs Loaded: " + str(nc1.getNumConfs()))
# Compute Energies of Conformations
print("Computing energies...")
_t1b = tm.time()
Ecmp1 = nc1.energy()
print("Energy computation complete. Time: " + "{:.4f}".format((tm.time() - _t1b) * 1000.0) + "ms")
# Compute Forces of Conformations
print("Compute forces...")
_t1b = tm.time()
F = nc1.force()
print("Force computation complete. Time: " + "{:.4f}".format((tm.time() - _t1b) * 1000.0) + "ms")
# Fn = np.array(nc1.computeNumericalForces(dr=0.0001))
n = smin
m = smax
Ecmp1 = gt.hatokcal * Ecmp1
Eact = gt.hatokcal * Eact
IDX = np.arange(0, Eact.shape[0], 1, dtype=float) * iscale + ishift
IDX = IDX
Eact = Eact
Ecmp1 = Ecmp1
Ecmp1 = Ecmp1 - Ecmp1.min()
Eact = Eact - Eact.min()
rmse1 = gt.calculaterootmeansqrerror(Eact, Ecmp1)
print("Spearman corr. 1: " + "{:.3f}".format(st.spearmanr(Ecmp1, Eact)[0]))
fig, axes = plt.subplots(nrows=2, ncols=2)
axes.flat[0].plot(IDX, Eact, "-", color="black", label="DFT", linewidth=6)
axes.flat[0].plot(IDX, Ecmp1, "--", color="red", label="ANI-1", linewidth=6)
# ax.plot(IDX, Eact, color='black', label='DFT', linewidth=3)
# ax.scatter(IDX, Eact, marker='o', color='black', linewidth=4)
th = axes.flat[0].set_title(title, fontsize=13)
th.set_position([0.5, 1.00])
# Set Limits
axes.flat[0].set_xlim([IDX.min(), IDX.max()])
# axes.flat[0].set_ylim([Eact.min()-1.0,Eact.max()+1.0])
axes.flat[0].set_ylabel("$\Delta$E calculated (kcal/mol)")
axes.flat[0].set_xlabel(xlabel)
axes.flat[0].legend(bbox_to_anchor=(0.2, 0.98), loc=2, borderaxespad=0.0, fontsize=12)
# print (Fn)
for i in range(0, 3):
Fq = F[:, :, i][:, atm]
# Fnq = Fn[:,i]
Faq = 1.8897259885789 * np.array(frc)[:, :, i][:, atm]
print(Fq)
print(Faq)
th = axes.flat[i + 1].set_title("Force for atom " + str(atm) + ": coordinate " + str(i), fontsize=13)
th.set_position([0.5, 1.00])
axes.flat[i + 1].set_xlim([IDX.min(), IDX.max()])
axes.flat[i + 1].plot(IDX, Faq, "-", color="black", label="DFT", linewidth=6)
# axes.flat[i+1].plot(IDX, Fnq, '-', color='blue', label='ANI Numerical',
# linewidth=6)
axes.flat[i + 1].plot(IDX, Fq, "--", color="red", label="ANI Analytical", linewidth=6)
axes.flat[i + 1].set_ylabel("Force (Ha/A)")
axes.flat[i + 1].set_xlabel(xlabel)
axes.flat[i + 1].legend(bbox_to_anchor=(0.2, 0.98), loc=2, borderaxespad=0.0, fontsize=12)
font = {"family": "Bitstream Vera Sans", "weight": "normal", "size": 12}
plt.rc("font", **font)
plt.show()
Ecmp3 = gt.hatokcal * Ecmp3
Ecmp4 = gt.hatokcal * Ecmp4
Ecmp5 = gt.hatokcal * Ecmp5
Eact = gt.hatokcal * Eact
Eotr = gt.hatokcal * Eotr
Emax = 400.0
Ecmp1 = setmaxE(Eact, Ecmp1, Emax)
Ecmp2 = setmaxE(Eact, Ecmp2, Emax)
Ecmp3 = setmaxE(Eact, Ecmp3, Emax)
Ecmp4 = setmaxE(Eact, Ecmp4, Emax)
Ecmp5 = setmaxE(Eact, Ecmp5, Emax)
Eotr = setmaxE(Eact, Eotr, Emax)
Eact = setmaxE(Eact, Eact, Emax)
rmse1 = gt.calculaterootmeansqrerror(Eact, Eotr)
rmse2 = gt.calculaterootmeansqrerror(Eact, Ecmp1)
rmse3 = gt.calculaterootmeansqrerror(Eact, Ecmp2)
rmse4 = gt.calculaterootmeansqrerror(Eact, Ecmp3)
rmse5 = gt.calculaterootmeansqrerror(Eact, Ecmp4)
rmse6 = gt.calculaterootmeansqrerror(Eact, Ecmp5)
mx = Eact.max()
mn = Eact.min()
#plt.scatter(IDX, Eact, marker='o' , color='black', linewidth=3)
print("Spearman corr. DFTB: " + "{:.3f}".format(st.spearmanr(Eotr, Eact)[0]))
print("Spearman corr. TGM 08: " +
"{:.3f}".format(st.spearmanr(Ecmp1, Eact)[0]))
#print ( "Spearman corr. TGM 07: " + "{:.3f}".format(st.spearmanr(Ecmp2,Eact)[0]) )
