def mypolyfit(x, y, order=1, verbose=1):
"""
coeff, yfit = mypolyfit(x,y,order=1, verbose=1)
"""
from numpy.lib.polynomial import polyfit, poly1d
coeffs = polyfit(x, y, order)
polyModel = poly1d(coeffs)
if verbose: print("Fit coeffs:", coeffs)
return coeffs, polyModel(x), polyModel
def C44_Calculator(self, EMT, PARAMETERS):
# Return 0 is used when the method is not desired to be used
# return 0
""" This method uses the given EMT calculator to calculate and return the value of the matrix element C44 for
a system of atoms of a given element type. The calculation is done by using that:
C44 = 1 / Volume * d^2/depsilon^2 (E_system) where epsilon is the displacement in one direction of the
system along one axis diveded by the highth of the system. """
# An atom object is created and the calculator attached
atoms = FaceCenteredCubic(size=(self.Size, self.Size, self.Size),
symbol=self.Element)
atoms.set_calculator(EMT)
# The volume of the sample is calculated
Vol = atoms.get_volume()
# The value of the relative displacement, epsilon, is set
epsilon = 1. / 1000
# The matrix used to change the unitcell by n*epsilon is initialized
LMM = numpy.array([[1, 0, -10 * epsilon], [0, 1, 0], [0, 0, 1]])
# The original unit cell is conserved
OCell = atoms.get_cell()
# The array for storing the energies is initialized
E_calc = numpy.zeros(20)
# The numerical value of C44 is calculated
for i in range(20):
# The new system cell based on the pertubation epsilon is set
atoms.set_cell(numpy.dot(OCell, LMM), scale_atoms=True)
# The energy of the system is calculated
E_calc[i] = atoms.get_potential_energy()
# The value of LMM is updated
LMM[0, 2] += epsilon
# A polynomial fit is made for the energy as a function of epsilon
# The displaced axis is defined
da = numpy.arange(-10, 10) * epsilon
# The fit is made
Poly = numpyfit.polyfit(da, E_calc, 2)
# Now C44 can be estimated from this fit by the second derivative of Poly = a * x^2 + b * x + c , multiplied
# with 1 / (2 * Volume) of system
C44 = 2. / Vol * Poly[0]
return C44
def G_reg1d(xx, yy, ww=None):
'''
计算斜率和截距
ww: weights
'''
rtn = []
ab = polyfit(xx, yy, 1, w=ww)
rtn.append(ab[0])
rtn.append(ab[1])
rtn.append(std(yy) / std(xx))
rtn.append(mean(yy) - rtn[2] * mean(xx))
r = corrcoef(xx, yy)
rr = r[0, 1] * r[0, 1]
rtn.append(rr)
return rtn
def G_reg1d(xx, yy, ww=None):
"""
description needed
ww: weights
"""
rtn = []
ab = polyfit(xx, yy, 1, w=ww)
rtn.append(ab[0])
rtn.append(ab[1])
rtn.append(std(yy) / std(xx))
rtn.append(mean(yy) - rtn[2] * mean(xx))
r = corrcoef(xx, yy)
rr = r[0, 1] * r[0, 1]
rtn.append(rr)
return rtn
def linearRegression(x, y, deg = 1, plotData = False):
'''returns the least square estimation of the linear regression of y = ax+b
as well as the plot'''
from numpy.lib.polynomial import polyfit
from matplotlib.pyplot import plot
from numpy.core.multiarray import arange
coef = polyfit(x, y, deg)
if plotData:
def poly(x):
result = 0
for i in range(len(coef)):
result += coef[i]*x**(len(coef)-i-1)
return result
plot(x, y, 'x')
xx = arange(min(x), max(x),(max(x)-min(x))/1000)
plot(xx, [poly(z) for z in xx])
return coef
def linearRegression(x, y, deg=1, plotData=False):
'''returns the least square estimation of the linear regression of y = ax+b
as well as the plot'''
from numpy.lib.polynomial import polyfit
from numpy.core.multiarray import arange
coef = polyfit(x, y, deg)
if plotData:
def poly(x):
result = 0
for i in range(len(coef)):
result += coef[i] * x**(len(coef) - i - 1)
return result
plt.plot(x, y, 'x')
xx = arange(min(x), max(x), (max(x) - min(x)) / 1000)
plt.plot(xx, [poly(z) for z in xx])
return coef
def list_derivative(f, x, n, o=3, p=5):
"""
Returns the nth derivative of f with respect to x, where zip(x,f) are
datapoints, using curve-fit polynomials of order o to p points.
"""
assert p % 2 == 1 # should be odd imo
assert len(f) >= p # won't work otherwise
derivative = []
for i in xrange(len(f)):
start = i - (p - 1) / 2
if start < 0:
start = 0
end = start + p
if end >= len(f):
start = -1 - p
end = -1
spline = polyfit(x[start:end], f[start:end], o)
derivative.append(polyval(polyder(spline, n), x[i]))
print("diff in list lengths: " + str(len(f) - len(derivative)))
return array(derivative)
def L12_BM_LC_Calculator(self, EMT, PARAMETERS):
# Return 0 is used when the method is not desired to be used
# return 0
""" The method L12_BM_LC_Calculator uses the EMT calculator to find the lattice constant which gives
the lowest possible energy for an alloy of two elements and from a polynomial fit of the energy as a
function of the lattice constant it calculates and returns the Bulk modulus, the volume of the system at
the lowest possible energy and this energy """
# Three values for the lattice constant, a0, are chosen. The highest and lowest value is chosen so that it
# is certain that these are to high/low respectively compared to the "correct" value. This is done by using the
# experimental value of the lattice constant, a0_exp, as the middle value and to high/low values are then:
# a0_exp +- a0_mod, a0_mod = a0_exp/10
a0_exp = ((PARAMETERS[self.Element[0]][1] +
3 * PARAMETERS[self.Element[1]][1]) * numpy.sqrt(2) * beta *
Bohr / 4)
a0_mod = a0_exp * 0.20
a0_guesses = numpy.array([a0_exp - a0_mod, a0_exp, a0_exp + a0_mod])
# An atoms object consisting of atoms of the chosen element is initialized
atoms = L1_2(size=(self.Size, self.Size, self.Size),
symbol=self.Element,
latticeconstant=a0_exp)
atoms.set_calculator(EMT)
# An identity matrix for the system is saved to a variable
IdentityMatrix = numpy.array([[1, 0, 0], [0, 1, 0], [0, 0, 1]])
# An array for the energy for the chosen guesses for a0 is initialized:
E_guesses = numpy.zeros(5)
# The energies are calculated
for i in range(3):
# Changes the lattice constant for the atoms object
atoms.set_cell(a0_guesses[i] * self.Size * IdentityMatrix,
scale_atoms=True)
# Calculates the energy of the system for the new lattice constant
E_guesses[i] = atoms.get_potential_energy()
# Bisection is used in order to find a small interval of the lattice constant for the minimum of the energy (an
# abitrary interval length is chosen), This is possible because we are certian from theory that there is only
# one minimum of the energy function in the interval of interest and thus we wont "fall" into a local minimum
# and stay there by accident.
while (a0_guesses[2] - a0_guesses[0]) >= self.Tol:
if min([E_guesses[0], E_guesses[2]]) == E_guesses[0]:
# A new guess for the lattice constant is introduced
a0_new_guess = 0.67 * a0_guesses[1] + 0.33 * a0_guesses[2]
# The energy for this new guess is calculated
atoms.set_cell(a0_new_guess * self.Size * IdentityMatrix,
scale_atoms=True)
E_new_guess = atoms.get_potential_energy()
# A check for changes in the energy minimum is made and the guesses for a0 and their corrosponding
# energies are ajusted.
if min(E_new_guess, min(E_guesses[0:3])) != E_new_guess:
a0_guesses[2] = a0_new_guess
E_guesses[2] = E_new_guess
else:
a0_guesses[0] = a0_guesses[1]
a0_guesses[1] = a0_new_guess
E_guesses[0] = E_guesses[1]
E_guesses[1] = E_new_guess
elif min([E_guesses[0], E_guesses[2]]) == E_guesses[2]:
# A new guess for the lattice constant is introduced
a0_new_guess = 0.33 * a0_guesses[0] + 0.67 * a0_guesses[1]
# The energy for this new guess is calculated
atoms.set_cell(a0_new_guess * self.Size * IdentityMatrix,
scale_atoms=True)
E_new_guess = atoms.get_potential_energy()
# A check for changes in the energy minimum is made and the guesses for a0 and their corrosponding
# energies are ajusted.
if min(E_new_guess, min(E_guesses[0:3])) != E_new_guess:
a0_guesses[0] = a0_new_guess
E_guesses[0] = E_new_guess
else:
a0_guesses[2] = a0_guesses[1]
a0_guesses[1] = a0_new_guess
E_guesses[2] = E_guesses[1]
E_guesses[1] = E_new_guess
# An estimate of the minimum energy can now be found from a second degree polynomial fit through the three
# current guesses for a0 and the corresponding values of the energy.
Poly = numpyfit.polyfit(a0_guesses, E_guesses[0:3], 2)
# The lattice constant corresponding to the lowest energy from the Polynomiel fit is found
a0 = -Poly[1] / (2 * Poly[0])
# Now five guesses for a0 and the corresponding energy are evaluated from and around the current a0.
a0_guesses = a0 * numpy.array([
1 - 2 * self.Tol / 5, 1 - self.Tol / 5, 1, 1 + self.Tol / 5,
1 + 2 * self.Tol / 5
])
for i in range(5):
# Changes the lattice constant for the atoms object
atoms.set_cell(a0_guesses[i] * self.Size * IdentityMatrix,
scale_atoms=True)
# Calculates the energy of the system for the new lattice constant
E_guesses[i] = atoms.get_potential_energy()
# The method EquationOfState is now used to find the Bulk modulus and the minimum energy for the system.
# The volume of the sample for the given a0_guesses
Vol = (self.Size * a0_guesses)**3
# The equilibrium volume, energy and bulk modulus are calculated
(Vol0, E0, B) = EquationOfState(Vol.tolist(), E_guesses.tolist()).fit()
return (Vol0, E0, B, Vol0**(1. / 3) / self.Size)
plt.title("Best Fit Line")
showPlot()
#Print s_square
error_var = squared_error(yVals, regLine) / len(yVals)
print "s_square= " + str(error_var)
print "RMSE= " + str((mean_squared_error(yVals, regLine, p))**0.5)
r_squared = coefficient_of_determination(yVals, regLine)
print "R_squared: " + str(r_squared)
F_val = calc_F(yVals, regLine, p)
print "F Statistics: " + str(F_val)
#trying Polynomial Fits
pl1 = polyfit(xVals, yVals, 1)
pl2 = polyfit(xVals, yVals, 2)
pl3 = polyfit(xVals, yVals, 3)
# print pl1
# print pl2
# print pl3
# plt.plot(xVals, yVals, 'o')
# plt.plot(xVals, np.polyval(pl1,xVals), 'r', linewidth=0.1)
# plt.plot(xVals, np.polyval(pl2,xVals), 'b', linewidth=0.5)
# plt.plot(xVals, np.polyval(pl3,xVals), 'g', linewidth=0.1)
# showPlot()
yfit = pl1[0] * xVals + pl1[1]
# yfit = pl2[0] * xVals + pl2[1]
# yfit = pl3[0] * xVals + pl3[1]
def analyze_olfaction_covariance(covariance, receptors):
''' Covariance: n x n covariance matrix.
Positions: list of n positions '''
positions = [pose.get_2d_position() for pose, sens in receptors] #@UnusedVariable
positions = array(positions).transpose().squeeze()
require_shape(square_shape(), covariance)
n = covariance.shape[0]
require_shape((2, n), positions)
distances = create_distance_matrix(positions)
correlation = cov2corr(covariance)
flat_distances = distances.reshape(n * n)
flat_correlation = correlation.reshape(n * n)
# let's fit a polynomial
deg = 4
poly = polyfit(flat_distances, flat_correlation, deg=deg)
knots = linspace(min(flat_distances), max(flat_distances), 2000)
poly_int = polyval(poly, knots)
poly_fder = polyder(poly)
fder = polyval(poly_fder, distances)
Ttheta = create_olfaction_Ttheta(positions, fder)
Tx, Ty = create_olfaction_Txy(positions, fder, distances)
report = Node('olfaction-theory')
report.data('flat_distances', flat_distances)
report.data('flat_correlation', flat_correlation)
with report.data_file('dist_vs_corr', 'image/png') as filename:
pylab.figure()
pylab.plot(flat_distances, flat_correlation, '.')
pylab.plot(knots, poly_int, 'r-')
pylab.xlabel('distance')
pylab.ylabel('correlation')
pylab.title('Correlation vs distance')
pylab.legend(['data', 'interpolation deg = %s' % deg])
pylab.savefig(filename)
pylab.close()
with report.data('fder', fder).data_file('fder', 'image/png') as filename:
pylab.figure()
pylab.plot(knots, polyval(poly_fder, knots), 'r-')
pylab.title('f der')
pylab.savefig(filename)
pylab.close()
report.data('distances', distances)
report.data('correlation', correlation)
report.data('covariance', covariance)
report.data('f', polyval(poly, distances))
f = report.figure(id='cor-vs-distnace', caption='Estimated kernels',
shape=(3, 3))
f.sub('dist_vs_corr')
f.sub('fder')
f.sub('f', display='scale')
f.sub('distances', display='scale')
f.sub('correlation', display='posneg')
f.sub('covariance', display='posneg')
T = numpy.zeros(shape=(3, Tx.shape[0], Tx.shape[1]))
T[0, :, :] = Tx
T[1, :, :] = Ty
T[2, :, :] = Ttheta
T_report = create_report_figure_tensors(T, report_id='tensors',
caption="Predicted learned tensors")
report.add_child(T_report)
return report
def redise(p0, x, y):
a, b = p0
return (y - (a * x + b))
data = sp.genfromtxt("web_traffic.tsv", "\t")
y = data[:, 1]
x = data[:, 0]
x_clean = x[np.where(y > -1)]
y_clean = y[np.where(y > -1)]
rs1 = leastsq(redise, [0, 0], args=(x_clean, y_clean))
rs2 = polyfit(x_clean, y_clean, 2, full=True)
rs5 = polyfit(x_clean, y_clean, 25, full=True)
y_predit1 = rs1[0][0] * x_clean + rs1[0][1]
y_predit2 = sp.poly1d(rs2[0])
y_predit5 = sp.poly1d(rs5[0])
print y_predit2
result = fsolve(y_predit2 - 10000, 0)
print result
pt.plot(x_clean, y_clean, '.')
pt.plot(x_clean, y_predit1, '-', label="predit1")
pt.plot(x_clean, y_predit2(x_clean), '-', label="predit2")
pt.plot(x_clean, y_predit5(x_clean), '-', label='predit5')
pt.legend()
pt.show()
