"""Example of nDimArrhenius Fit."""
from Stoner import Data
import Stoner.Fit as SF
from numpy import linspace
from numpy.random import normal
#Make some data
T = linspace(50, 500, 101)
R = SF.nDimArrhenius(T + normal(size=len(T), scale=5.0, loc=1.0), 1E6, 0.5, 2)
d = Data(T, R, setas="xy", column_headers=["T", "Rate"])
#Curve_fit on its own
d.curve_fit(SF.nDimArrhenius,
p0=[1E6, 0.5, 2],
result=True,
header="curve_fit")
d.setas = "xyy"
d.plot()
d.annotate_fit(SF.nDimArrhenius, x=150, y=6E5)
# lmfit using lmfit guesses
fit = SF.NDimArrhenius()
p0 = fit.guess(R, x=T)
d.lmfit(fit, p0=p0, result=True, header="lmfit")
d.setas = "x..y"
d.plot()
d.annotate_fit(SF.NDimArrhenius, x=150, y=3.5E5, prefix="NDimArrhenius")
d.title = "n-D Arrhenius Test Fit"
d.ylabel = "Rate"
d.xlabel = "Temperature (K)"
d.plot(fmt="m-", label="Polyfit")
d.text(
-9,
450,
"Polynominal co-efficients\n{}".format(d["2nd-order polyfit coefficients"]),
fontdict={"size": "x-small", "color": "magenta"},
)
d.setas = "xy"
d.curve_fit(SF.quadratic, result=True, header="Curve-fit")
d.setas = "x...y"
d.plot(fmt="b-", label="curve-fit")
d.annotate_fit(
SF.quadratic,
prefix="quadratic",
x=0.2,
y=0.65,
fontdict={"size": "x-small", "color": "blue"},
)
d.setas = "xy"
fit = SF.Quadratic()
p0 = fit.guess(y, x=x)
d.lmfit(SF.Quadratic, p0=p0, result=True, header="lmfit")
d.setas = "x...y"
d.plot(fmt="g-", label="lmfit")
d.annotate_fit(
SF.Quadratic,
prefix="Quadratic",
x=0.65,
# Curve_fit on its own
d.curve_fit(vftEquation, p0=params, result=True, header="curve_fit")
# lmfit uses some guesses
p0 = params
d.lmfit(VFTEquation, result=True, header="lmfit")
# Plot these results too
d.setas = "x..yy"
d.plot(fmt=["b-", "g-"])
# Annotate the graph
d.annotate_fit(
vftEquation,
x=0.25,
y=0.35,
fontdict={
"size": "x-small",
"color": "blue"
},
mode="eng",
)
d.annotate_fit(
VFTEquation,
x=0.5,
y=0.35,
prefix="VFTEquation",
fontdict={
"size": "x-small",
"color": "green"
},
mode="eng",
)
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Fit Ic(B) to Airy function.
"""
from Stoner import Data, __home__
from Stoner.Fit import Ic_B_Airy
import os
os.chdir(os.path.join(__home__, "..", "doc", "samples", "Fitting"))
data = Data("data/Ic_B.txt", setas={"x": "Magnet Output", "y": "Ic"})
data.lmfit(Ic_B_Airy, result=True, header="Fit")
data.setas = {"x": "Magnet Output", "y": ["Ic", "Fit"]}
data.plot(fmt=["r+", "b-"])
data.annotate_fit(Ic_B_Airy, mode="eng", x=0.6, y=0.5, fontsize="small")
data.title = "Critical current vs Field for $4\mu m$ junction"
data.xlabel = r"Magnetic Field $\mu_0H (\mathrm{T})$"
data.ylabel = r"Critical Current $I_c (A)$"
data.del_rows(isnan(data.y))
#Normalise data on y axis between +/- 1
data.normalise(base=(-1.,1.), replace=True)
#Swap x and y axes around so that R is x and T is y
data=~data
#Curve fit a straight line, using only the central 90% of the resistance transition
data.curve_fit(linear,bounds=lambda x,r:-threshold<x<threshold,result=True,p0=[7.0,0.0]) #result=True to record fit into metadata
#Plot the results
data.setas[-1]="y"
data.subplot(1,len(r_cols),i+1)
data.plot(fmt=["k.","r-"])
data.annotate_fit(linear,x=-1.,y=7.3c,fontsize="small")
data.title="Ramp {}".format(data[iterator][0])
row.extend([data["linear:intercept"],data["linear:intercept err"]])
data.tight_layout()
result+=np.array(row)
result.column_headers=["Ramp","Sample 4 R","dR","Sample 7 R","dR"]
result.setas="xyeye"
result.plot(fmt=["k.","r."])
"""Example of PowerLaw Fit.""" from Stoner import Data import Stoner.Fit as SF from numpy import linspace from numpy.random import normal # Make some data T = linspace(50, 500, 101) R = SF.powerLaw(T, 1e-2, 0.6666666) * normal(size=len(T), scale=0.1, loc=1.0) d = Data(T, R, setas="xy", column_headers=["T", "Rate"]) # Curve_fit on its own d.curve_fit(SF.powerLaw, p0=[1, 0.5], result=True, header="curve_fit") d.setas = "xyy" d.plot(fmt=["r.", "b-"]) d.annotate_fit(SF.powerLaw, x=0.5, y=0.25) # lmfit using lmfit guesses fit = SF.PowerLaw() p0 = fit.guess(R, x=T) d.lmfit(fit, p0=p0, result=True, header="lmfit") d.setas = "x..y" d.plot(fmt="g-") d.annotate_fit(SF.PowerLaw, x=0.5, y=0.05, prefix="PowerLaw") d.title = "Powerlaw Test Fit" d.ylabel = "Rate" d.xlabel = "Temperature (K)"
from Stoner import Data
import Stoner.Fit as SF
from numpy import linspace,ones_like
from numpy.random import normal
#Make some data
V=linspace(-4,4,1000)
I=SF.fowlerNordheim(V,2500,3.2,15.0)+normal(size=len(V),scale=10E-6)
dI=ones_like(V)*10E-6
d=Data(V,I,dI,setas="xye",column_headers=["Bias","Current","Noise"])
d.curve_fit(SF.fowlerNordheim,p0=[2500,3.2,15.0],result=True,header="curve_fit")
d.setas="xyey"
d.plot(fmt=["r.","b-"])
d.annotate_fit(SF.fowlerNordheim,x=0,y=10,prefix="fowlerNordheim",fontdict={"size":"x-small"})
d.setas="xye"
fit=SF.FowlerNordheim()
p0=[2500,5.2,15.0]
p0=fit.guess(I,x=V)
for p,v,mi,mx in zip(["A","phi","d"],[2500,3.2,15.0],[100,1,5],[1E4,20.0,30.0]):
p0[p].value=v
p0[p].bounds=[mi,mx]
d.lmfit(SF.FowlerNordheim,p0=p0,result=True,header="lmfit")
d.setas="x...y"
d.plot()
d.annotate_fit(fit,x=-3,y=-60,prefix="FowlerNordheim",fontdict={"size":"x-small"})
d.ylabel="Current"
d.title="Fowler-Nordheim Model test"
G = SF.fluchsSondheimer(B, *params) + normal(size=len(B), scale=5e-5)
dG = ones_like(B) * 5e-5
d = Data(
B,
G,
dG,
setas="xye",
column_headers=["Thickness (nm)", "Conductance", "dConductance"],
)
d.curve_fit(SF.fluchsSondheimer, p0=params, result=True, header="curve_fit")
d.setas = "xye"
d.lmfit(SF.FluchsSondheimer, p0=params, result=True, header="lmfit")
d.setas = "xyeyy"
d.plot(fmt=["r.", "b-", "g-"])
d.annotate_fit(
SF.fluchsSondheimer, x=0.2, y=0.6, fontdict={"size": "x-small", "color": "blue"}
)
d.annotate_fit(
SF.FluchsSondheimer,
x=0.2,
y=0.4,
fontdict={"size": "x-small", "color": "green"},
prefix="FluchsSondheimer",
)
d.title = "Fluchs-Sondheimer Fit"
d.tight_layout()
data.del_rows(isnan(data.y))
#Normalise data on y axis between +/- 1
data.normalise(base=(-1.,1.), replace=True)
#Swap x and y axes around so that R is x and T is y
data=~data
#Curve fit a straight line, using only the central 90% of the resistance transition
data.curve_fit(linear,bounds=lambda x,r:-threshold<x<threshold,result=True,p0=[7.0,0.0]) #result=True to record fit into metadata
#Plot the results
data.setas[-1]="y"
data.subplot(1,len(r_cols),i+1)
data.plot(fmt=["k.","r-"])
data.annotate_fit(linear,x=-1.,y=7.3c,fontsize="small")
data.title="Ramp {}".format(data[iterator][0])
row.extend([data["linear:intercept"],data["linear:intercept err"]])
data.tight_layout()
result+=np.array(row)
result.column_headers=["Ramp","Sample 4 R","dR","Sample 7 R","dR"]
result.setas="xyeye"
result.plot(fmt=["k.","r."])
@simple_model.guesser
def guess_vals(y, x=None):
"""Should guess parameter values really!"""
m = (y.max() - y.min()) / (x[y.argmax()] - x[y.argmin()])
c = x.mean() * m - y.mean() # return one value per parameter
return [m, c]
# Add a function to sry vonstraints on parameters (optional)
@simple_model.hinter
def hint_parameters():
"""Five some hints about the parameter."""
return {"m": {"max": 10.0, "min": 0.0}, "c": {"max": 5.0, "min": -5.0}}
# Create some x,y data
x = linspace(0, 10, 101)
y = 4.5 * x - 2.3 + normal(scale=0.4, size=len(x))
# Make The Data object
d = Data(x, y, setas="xy", column_headers=["X", "Y"])
# Do the fit
d.lmfit(simple_model, result=True)
# Plot the result
d.setas = "xyy"
d.plot(fmt=["r+", "b-"])
d.title = "Simple Model Fit"
d.annotate_fit(simple_model, x=0.05, y=0.5)
# Curve_fit on its own
d.curve_fit(vftEquation, p0=params, result=True, header="curve_fit")
# lmfit uses some guesses
p0 = params
d.lmfit(VFTEquation, p0=p0, result=True, header="lmfit")
# Plot these results too
d.setas = "x..yy"
d.plot(fmt=["b-", "g-"])
# Annotate the graph
d.annotate_fit(
vftEquation,
x=0.25,
y=0.35,
fontdict={"size": "x-small", "color": "blue"},
mode="eng",
)
d.annotate_fit(
VFTEquation,
x=0.5,
y=0.35,
prefix="VFTEquation",
fontdict={"size": "x-small", "color": "green"},
mode="eng",
)
# reset the columns for the fit
d.setas = "xye.."
# Now do the odr fit (will overwrite lmfit's metadata)
)
d.curve_fit(SF.kittelEquation, p0=copy(params), result=True, header="curve_fit")
fit = SF.KittelEquation()
p0 = fit.guess(G, x=B)
d.lmfit(fit, p0=p0, result=True, header="lmfit")
d.setas = "xyeyy"
d.plot(fmt=["r.", "b-", "g-"])
d.annotate_fit(
SF.kittelEquation,
x=0.5,
y=0.25,
fontdict={"size": "x-small", "color": "blue"},
mode="eng",
)
d.annotate_fit(
SF.KittelEquation,
x=0.5,
y=0.05,
fontdict={"size": "x-small", "color": "green"},
prefix="KittelEquation",
mode="eng",
)
d.title = "Kittel Fit"
d.fig.gca().xaxis.set_major_formatter(TexEngFormatter())
d.fig.gca().yaxis.set_major_formatter(TexEngFormatter())
d.tight_layout()
from numpy.random import normal
# Make some data
T = linspace(200, 350, 101)
R = SF.arrhenius(T + normal(size=len(T), scale=3.0, loc=0.0), 1e6, 0.5)
E = 10 ** ceil(log10(abs(R - SF.arrhenius(T, 1e6, 0.5))))
d = Data(T, R, E, setas="xye", column_headers=["T", "Rate"])
# Curve_fit on its own
d.curve_fit(SF.arrhenius, p0=(1e6, 0.5), result=True, header="curve_fit")
d.setas = "xyey"
d.plot(fmt=["r.", "b-"])
d.annotate_fit(
SF.arrhenius,
x=0.5,
y=0.5,
mode="eng",
fontdict={"size": "x-small", "color": "blue"},
)
# lmfit using lmfit guesses
fit = SF.Arrhenius()
d.setas = "xye"
d.lmfit(fit, result=True, header="lmfit")
d.setas = "x...y"
d.plot(fmt="g-")
d.annotate_fit(
SF.Arrhenius,
x=0.5,
y=0.35,
prefix="Arrhenius",
result.data[:, c] = (resfldr[1][:, c] + resfldr[0][:, c]) / 2.0
for c in [1, 3, 5, 7]:
result.data[:, c] = gmean((resfldr[0][:, c], resfldr[1][:, c]), axis=0)
# Doing the Kittel fit with an orthogonal distance regression as we have x errors not y errors
p0 = [2, 200e3, 10e3] # Some sensible guesses
result.lmfit(
Inverse_Kittel, p0=p0, result=True, header="Kittel Fit", output="report"
)
result.setas[-1] = "y"
result.template.yformatter = TexEngFormatter
result.template.xformatter = TexEngFormatter
result.labels = None
result.figure(figsize=(6, 8))
result.subplot(211)
result.plot(fmt=["r.", "b-"])
result.annotate_fit(Inverse_Kittel, x=7e9, y=1e5, fontdict={"size": 8})
result.ylabel = "$H_{res} \\mathrm{(Am^{-1})}$"
result.title = "Inverse Kittel Fit"
# Get alpha
result.subplot(212)
result.setas(y="Delta_H", e="Delta_H.stderr", x="Freq")
result.y /= mu_0
result.e /= mu_0
result.lmfit(Linear, result=True, header="Width", output="report")
result.setas[-1] = "y"
result.plot(fmt=["r.", "b-"])
result.annotate_fit(Linear, x=5.5e9, y=2.8e3, fontdict={"size": 8})
setas="xye",
column_headers=["Field $Oe$", r"$\nu (Hz)$", r"\delta $\nu (Hz)$"])
d.curve_fit(SF.kittelEquation,
p0=copy(params),
result=True,
header="curve_fit")
fit = SF.KittelEquation()
p0 = fit.guess(G, x=B)
d.lmfit(fit, p0=p0, result=True, header="lmfit")
d.setas = "xyeyy"
d.plot(fmt=["r.", "b-", "g-"])
d.annotate_fit(SF.kittelEquation,
x=4000,
y=2E8,
fontdict={"size": "x-small"},
mode="eng")
d.annotate_fit(SF.KittelEquation,
x=20000,
y=2E8,
fontdict={"size": "x-small"},
prefix="KittelEquation",
mode="eng")
d.title = "Kittel Fit"
d.fig.gca().xaxis.set_major_formatter(TexEngFormatter())
d.fig.gca().yaxis.set_major_formatter(TexEngFormatter())
d.tight_layout()
"""USe curve_fit to fit a straight line."""
from Stoner import Data
def linear(x, m, c):
"""Straight line function."""
return m * x + c
d = Data("curve_fit_data.dat", setas="xye")
d.plot(fmt="ro")
fit = d.curve_fit(linear, result=True, replace=False, header="Fit", output="report")
d.setas = "x..y"
d.plot(fmt="b-")
d.annotate_fit(linear)
d.draw()
print(fit.fit_report())
from numpy import linspace, ones_like
from numpy.random import normal
# Make some data
V = linspace(-10, 10, 1000)
I = SF.bdr(V, 2.5, 3.2, 0.3, 15.0, 1.0) + normal(size=len(V), scale=1.0e-3)
dI = ones_like(V) * 1.0e-3
# Curve fit
d = Data(V, I, dI, setas="xye", column_headers=["Bias", "Current", "Noise"])
d.curve_fit(SF.bdr, p0=[2.5, 3.2, 0.3, 15.0, 1.0], result=True, header="curve_fit")
d.setas = "xyey"
d.plot(fmt=["r.", "b-"])
d.annotate_fit(
SF.bdr, x=0.6, y=0.05, prefix="bdr", fontdict={"size": "x-small", "color": "blue"}
)
# lmfit
d.setas = "xy"
fit = SF.BDR(missing="drop")
p0 = fit.guess(I, x=V)
for p, v, mi, mx in zip(
["A", "phi", "dphi", "d", "mass"],
[2.500, 3.2, 0.3, 15.0, 1.0],
[0.100, 1.0, 0.05, 5.0, 0.5],
[10, 10.0, 2.0, 30.0, 5.0],
):
p0[p].value = v
p0[p].bounds = [mi, mx]
d.lmfit(fit, p0=p0, result=True, header="lmfit")
B = linspace(-8, 8, 201)
params = [1e-3, 2.0, 0.25, 1.4]
G = SF.wlfit(B, *params) + normal(size=len(B), scale=5e-7)
dG = ones_like(B) * 5e-7
d = Data(
B,
G,
dG,
setas="xye",
column_headers=["Field $\\mu_0H (T)$", "Conductance", "dConductance"],
)
d.curve_fit(SF.wlfit, p0=copy(params), result=True, header="curve_fit")
d.setas = "xye"
d.lmfit(SF.WLfit, p0=copy(params), result=True, header="lmfit")
d.setas = "xyeyy"
d.plot(fmt=["r.", "b-", "g-"])
d.annotate_fit(SF.wlfit, x=0.05, y=0.75, fontdict={"size": "x-small", "color": "blue"})
d.annotate_fit(
SF.WLfit,
x=0.05,
y=0.5,
fontdict={"size": "x-small", "color": "green"},
prefix="WLfit",
)
d.title = "Weak Localisation Fit"
d.tight_layout()
from numpy.random import normal
# Make some data
T = linspace(200, 350, 101)
R = SF.arrhenius(T + normal(size=len(T), scale=3.0, loc=0.0), 1e6, 0.5)
E = 10 ** ceil(log10(np_abs(R - SF.arrhenius(T, 1e6, 0.5))))
d = Data(T, R, E, setas="xye", column_headers=["T", "Rate"])
# Curve_fit on its own
d.curve_fit(SF.arrhenius, p0=(1e6, 0.5), result=True, header="curve_fit")
d.setas = "xyey"
d.plot(fmt=["r.", "b-"])
d.annotate_fit(
SF.arrhenius,
x=0.5,
y=0.5,
mode="eng",
fontdict={"size": "x-small", "color": "blue"},
)
# lmfit using lmfit guesses
fit = SF.Arrhenius()
d.setas = "xye"
d.lmfit(fit, result=True, header="lmfit")
d.setas = "x...y"
d.plot(fmt="g-")
d.annotate_fit(
SF.Arrhenius,
x=0.5,
y=0.35,
prefix="Arrhenius",
from numpy.random import normal
# Make some data
V = linspace(-4, 4, 1000)
I = SF.fowlerNordheim(V, 2500, 3.2, 15.0) + normal(size=len(V), scale=1e-6)
dI = ones_like(V) * 10e-6
d = Data(V, I, dI, setas="xye", column_headers=["Bias", "Current", "Noise"])
d.curve_fit(SF.fowlerNordheim, p0=[2500, 3.2, 15.0], result=True, header="curve_fit")
d.setas = "xyey"
d.plot(fmt=["r.", "b-"])
d.annotate_fit(
SF.fowlerNordheim,
x=0.2,
y=0.6,
prefix="fowlerNordheim",
fontdict={"size": "x-small", "color": "blue"},
)
d.setas = "xye"
fit = SF.FowlerNordheim()
p0 = [2500, 5.2, 15.0]
p0 = fit.guess(I, x=V)
for p, v, mi, mx in zip(
["A", "phi", "d"], [2500, 3.2, 15.0], [100, 1, 5], [1e4, 20.0, 30.0]
):
p0[p].value = v
p0[p].bounds = [mi, mx]
d.lmfit(SF.FowlerNordheim, p0=p0, result=True, header="lmfit")
d.setas = "x...y"
"""Example of using lmfit to do a bounded fit."""
from Stoner import Data
from Stoner.Fit import StretchedExp
#Load dat and plot
d = Data("lmfit_data.txt", setas="xy")
# Do the fit
d.lmfit(StretchedExp, result=True, header="Fit", prefix="")
# plot
d.setas = "xyy"
d.plot(fmt=["+", "-"])
# Make apretty label using Stoner.Util methods
text = "$y=A\\exp\\left[\\left(-\\frac{x}{x_0}\\right)^\\beta\\right]$\n"
text += d.annotate_fit(StretchedExp, text_only=True)
d.text(6, 4E4, text)
#Adjust layout NB pass-through method to pyplot used here
d.tight_layout()
"""USe curve_fit to fit a straight line."""
from Stoner import Data
def linear(x, m, c):
"""Straight line function."""
return m * x + c
d = Data("curve_fit_data.dat", setas="xye")
d.plot(fmt="ro")
popt, pcov = d.curve_fit(linear, result=True, replace=False, header="Fit")
d.setas = "x..y"
d.plot(fmt="b-")
d.annotate_fit(linear)
d.draw()
from numpy.random import normal
# Make some data
V = linspace(-4, 4, 1001)
I = SF.simmons(V, 2500, 5.2, 15.0) + normal(size=len(V), scale=100e-9)
dI = ones_like(V) * 100e-9
d = Data(V, I, dI, setas="xye", column_headers=["Bias", "Current", "Noise"])
d.curve_fit(SF.simmons, p0=[2500, 5.2, 15.0], result=True, header="curve_fit")
d.setas = "xyey"
d.plot(fmt=["r.", "b-"])
d.annotate_fit(
SF.simmons,
x=0.25,
y=0.25,
prefix="simmons",
fontdict={"size": "x-small", "color": "blue"},
)
d.setas = "xye"
fit = SF.Simmons()
p0 = [2500, 5.2, 15.0]
d.lmfit(SF.Simmons, p0=p0, result=True, header="lmfit")
d.setas = "x...y"
d.plot(fmt="g-")
d.annotate_fit(
fit,
x=0.65,
y=0.25,
prefix="Simmons",
# Make some data
V = linspace(-4, 4, 1001)
I = SF.simmons(V, 2500, 5.2, 15.0) + normal(size=len(V), scale=100e-9)
dI = ones_like(V) * 100e-9
d = Data(V, I, dI, setas="xye", column_headers=["Bias", "Current", "Noise"])
d.curve_fit(SF.simmons, p0=[2500, 5.2, 15.0], result=True, header="curve_fit")
d.setas = "xyey"
d.plot(fmt=["r.", "b-"])
d.annotate_fit(
SF.simmons,
x=0.25,
y=0.25,
prefix="simmons",
fontdict={
"size": "x-small",
"color": "blue"
},
)
d.setas = "xye"
fit = SF.Simmons()
p0 = [2500, 5.2, 15.0]
d.lmfit(SF.Simmons, p0=p0, result=True, header="lmfit")
d.setas = "x...y"
d.plot(fmt="g-")
d.annotate_fit(
fit,
x=0.65,
y=0.25,
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Fit IV data to various RSJ models
"""
from Stoner import Data, __home__
from Stoner.Fit import RSJ_Noiseless, RSJ_Simple
import os
os.chdir(os.path.join(__home__, "..", "doc", "samples", "Fitting"))
data = Data("data/IV.txt", setas={"x": "Current", "y": "Voltage"})
# Fit data with both versions of the RSJ model
data.lmfit(RSJ_Simple, result=True, header="Simple", prefix="simple")
data.lmfit(RSJ_Noiseless, result=True, header="Noiseless", prefix="noiseless")
# Set column assignments and plot the data and fits
data.setas = {"x": "Current", "y": ["Voltage", "Simple", "Noiseless"]}
data.plot(fmt=["r+", "b-", "g-"])
# Annotate fits
data.annotate_fit(
RSJ_Simple, prefix="simple", mode="eng", x=0.15, y=0.1, fontsize="small"
)
data.annotate_fit(
RSJ_Noiseless, prefix="noiseless", mode="eng", x=0.55, y=0.1, fontsize="small"
)
import Stoner.Fit as SF
from numpy import linspace
from numpy.random import normal
# Make some data
T = linspace(200, 350, 101)
R = SF.modArrhenius(T, 1e6, 0.5, 1.5) * normal(
scale=0.00005, loc=1.0, size=len(T))
d = Data(T, R, setas="xy", column_headers=["T", "Rate"])
# Curve_fit on its own
d.curve_fit(SF.modArrhenius,
p0=[1e6, 0.5, 1.5],
result=True,
header="curve_fit")
d.setas = "xyy"
d.plot(fmt=["r.", "b-"])
d.annotate_fit(SF.modArrhenius, x=0.2, y=0.5)
# lmfit using lmfit guesses
fit = SF.ModArrhenius()
p0 = [1e6, 0.5, 1.5]
d.lmfit(fit, p0=p0, result=True, header="lmfit")
d.setas = "x..y"
d.plot()
d.annotate_fit(SF.ModArrhenius, x=0.2, y=0.25, prefix="ModArrhenius")
d.title = "Modified Arrhenius Test Fit"
d.ylabel = "Rate"
d.xlabel = "Temperature (K)"
"""Example of nDimArrhenius Fit.""" from Stoner import Data import Stoner.Fit as SF from numpy import linspace from numpy.random import normal # Make some data T = linspace(50, 500, 101) R = SF.nDimArrhenius(T + normal(size=len(T), scale=5.0, loc=1.0), 1e6, 0.5, 2) d = Data(T, R, setas="xy", column_headers=["T", "Rate"]) # Curve_fit on its own d.curve_fit(SF.nDimArrhenius, p0=[1e6, 0.5, 2], result=True, header="curve_fit") d.setas = "xyy" d.plot(fmt=["r.", "b-"]) d.annotate_fit(SF.nDimArrhenius, x=0.25, y=0.3) # lmfit using lmfit guesses fit = SF.NDimArrhenius() p0 = fit.guess(R, x=T) d.lmfit(fit, p0=p0, result=True, header="lmfit") d.setas = "x..y" d.plot(fmt="g-") d.annotate_fit(SF.NDimArrhenius, x=0.25, y=0.05, prefix="NDimArrhenius") d.title = "n-D Arrhenius Test Fit" d.ylabel = "Rate" d.xlabel = "Temperature (K)"
d.curve_fit(func, p0=copy(params)[0:2], result=True, header="curve_fit")
d.setas = "xye"
fit = SF.Langevin()
p0 = fit.guess(G, x=B)
for p, v in zip(p0, params):
p0[p].set(v)
p0[p].max = v * 5
p0[p].min = 0
p0[p].vary = p != "T"
d.lmfit(fit, p0=p0, result=True, header="lmfit")
d.setas = "xyeyy"
d.plot(fmt=["r.", "b-", "g-"])
d.annotate_fit(
func, x=0.1, y=0.5, fontdict={"size": "x-small", "color": "blue"}, mode="eng"
)
d.annotate_fit(
SF.Langevin,
x=0.1,
y=0.25,
fontdict={"size": "x-small", "color": "green"},
prefix="Langevin",
mode="eng",
)
d.title = "langevin Fit"
setas="xye",
column_headers=["Thickness (nm)", "Conductance", "dConductance"],
)
d.curve_fit(SF.fluchsSondheimer, p0=params, result=True, header="curve_fit")
d.setas = "xye"
d.lmfit(SF.FluchsSondheimer, p0=params, result=True, header="lmfit")
d.setas = "xyeyy"
d.plot(fmt=["r.", "b-", "g-"])
d.annotate_fit(SF.fluchsSondheimer,
x=0.2,
y=0.6,
fontdict={
"size": "x-small",
"color": "blue"
})
d.annotate_fit(
SF.FluchsSondheimer,
x=0.2,
y=0.4,
fontdict={
"size": "x-small",
"color": "green"
},
prefix="FluchsSondheimer",
)
d.title = "Fluchs-Sondheimer Fit"
d.tight_layout()
import Stoner.Fit as SF
from numpy import linspace,ones_like
from numpy.random import normal
#Make some data
V=linspace(-10,10,1000)
I=SF.bdr(V,2.5,3.2,0.3,15.0,1.0)+normal(size=len(V),scale=1.0E-3)
dI=ones_like(V)*1.0E-3
#Curve fit
d=Data(V,I,dI,setas="xye",column_headers=["Bias","Current","Noise"])
d.curve_fit(SF.bdr,p0=[2.5,3.2,0.3,15.0,1.0],result=True,header="curve_fit")
d.setas="xyey"
d.plot(fmt=["r.","b-"])
d.annotate_fit(SF.bdr,x=-4,y=1.0,prefix="bdr",fontdict={"size":"x-small"})
#lmfit
d.setas="xy"
fit=SF.BDR(missing="drop")
p0=fit.guess(I,x=V)
for p,v,mi,mx in zip(["A","phi","dphi","d","mass"],[2.500,3.2,0.3,15.0,1.0],[0.100,1.0,0.05,5.0,0.5],[10,10.0,2.0,30.0,5.0]):
p0[p].value=v
p0[p].bounds=[mi,mx]
d.lmfit(fit,p0=p0,result=True,header="lmfit")
d.setas="x...y"
d.plot()
d.annotate_fit(fit,x=-4,y=-20.0,prefix="BDR",fontdict={"size":"x-small"})
d.ylabel="Current"
d.title="BDR Model test"
"""Example of using lmfit to do a bounded fit."""
from Stoner import Data
from Stoner.Fit import StretchedExp
# Load dat and plot
d = Data("lmfit_data.txt", setas="xy")
# Do the fit
d.lmfit(StretchedExp, result=True, header="Fit", prefix="")
# plot
d.setas = "xyy"
d.plot(fmt=["+", "-"])
# Make apretty label using Stoner.Util methods
text = r"$y=A e^{-\left(\frac{x}{x_0}\right)^\beta}$" + "\n"
text += d.annotate_fit(StretchedExp, text_only=True, prefix="")
d.text(6, 4e4, text)
@simple_model.guesser
def guess_vals(y, x=None):
"""Should guess parameter values really!"""
m = (y.max() - y.min()) / (x[y.argmax()] - x[y.argmin()])
c = x.mean() * m - y.mean() # return one value per parameter
return [m, c]
# Add a function to sry vonstraints on parameters (optional)
@simple_model.hinter
def hint_parameters():
"""Five some hints about the parameter."""
return {"m": {"max": 10.0, "min": 0.0}, "c": {"max": 5.0, "min": -5.0}}
# Create some x,y data
x = linspace(0, 10, 101)
y = 4.5 * x - 2.3 + normal(scale=0.4, size=len(x))
# Make The Data object
d = Data(x, y, setas="xy", column_headers=["X", "Y"])
# Do the fit
d.lmfit(simple_model, result=True)
# Plot the result
d.setas = "xyy"
d.plot(fmt=["r+", "b-"])
d.title = "Simple Model Fit"
d.annotate_fit(simple_model, x=0.05, y=0.5)
func,
result=True,
header="Fit",
A=1,
B=1,
C=1,
prefix="Model",
residuals=True,
)
# Reset labels
d.labels = []
# Make nice two panel plot layout
ax = d.subplot2grid((3, 1), (2, 0))
d.setas = "x..y"
d.plot(fmt="g+")
d.title = ""
ax = d.subplot2grid((3, 1), (0, 0), rowspan=2)
d.setas = "xyy"
d.plot(fmt=["r.", "b-"])
d.xticklabels = [[]]
d.xlabel = ""
# Annotate plot with fitting parameters
d.annotate_fit(func, prefix="Model", x=0.7, y=0.3, fontdict={"size": "x-small"})
text = r"$y=A+Be^{-x/C}$" + "\n\n"
d.text(7.2, 3.9, text, fontdict={"size": "x-small"})
d.title = u"Differential Evolution Fit"
A=1,
B=1,
C=1,
prefix="Model",
residuals=True)
#Reset labels
d.labels = []
# Make nice two panel plot layout
ax = d.subplot2grid((3, 1), (2, 0))
d.setas = "x..y"
d.plot(fmt="g+")
d.title = ""
ax = d.subplot2grid((3, 1), (0, 0), rowspan=2)
d.setas = "xyy"
d.plot(fmt=["ro", "b-"])
d.xticklabels = [[]]
d.xlabel = ""
# Annotate plot with fitting parameters
d.annotate_fit(func,
prefix="Model",
x=7.2,
y=3.45,
fontdict={"size": "x-small"})
text = r"$y=A+Be^{-x/C}$" + "\n\n"
d.text(7.2, 3.9, text, fontdict={"size": "x-small"})
d.title = u"Orthogonal Distance Regression Fit"
from numpy import linspace, ones_like
from numpy.random import normal
from copy import deepcopy
T = linspace(4.2, 300, 101)
params = [265, 65, 1.0, 5]
params2 = deepcopy(params)
G = SF.blochGrueneisen(T, *params) + normal(size=len(T), scale=5E-5)
dG = ones_like(T) * 5E-5
d = Data(T,
G,
dG,
setas="xye",
column_headers=["Temperature (K)", "Resistivity", "dR"])
d.curve_fit(SF.blochGrueneisen, p0=params, result=True, header="curve_fit")
d.setas = "xy"
d.lmfit(SF.BlochGrueneisen, p0=params2, result=True, header="lmfit")
d.setas = "xyeyy"
d.plot(fmt=["r.", "b-", "g-"])
d.annotate_fit(SF.blochGrueneisen, x=20, y=65.05, fontdict={"size": "x-small"})
d.annotate_fit(SF.BlochGrueneisen,
x=100,
y=65,
fontdict={"size": "x-small"},
prefix="BlochGrueneisen")
d.title = "Bloch-Grueneisen Fit"
d.tight_layout()
d = Data(T, R, dR, setas="xye", column_headers=["T", "Rate"])
#Curve_fit on its own
d.curve_fit(SF.vftEquation, p0=params, result=True, header="curve_fit")
# lmfit using lmfit guesses
fit = SF.VFTEquation()
p0 = params
d.lmfit(fit, p0=p0, result=True, header="lmfit")
d.setas = "xyeyyy"
d.plot(fmt=["k+", "r-", "b-"])
d.yscale = "log"
d.ylim = (1E-43, 1)
d.annotate_fit(SF.vftEquation,
x=270,
y=1E-27,
fontdict={"size": "x-small"},
mode="eng")
d.annotate_fit(SF.VFTEquation,
x=240,
y=1E-40,
prefix="VFTEquation",
fontdict={"size": "x-small"},
mode="eng")
d.title = "VFT Equation Test Fit"
d.ylabel = "Rate"
d.xlabel = "Temperature (K)"
d.plot(fmt="ro") # plot our data
func = lambda x, A, B, C: A + B * exp(-x / C)
# Do the fitting and plot the result
fit = d.differential_evolution(
func, result=True, header="Fit", A=1, B=1, C=1, prefix="Model", residuals=True
)
# Reset labels
d.labels = []
# Make nice two panel plot layout
ax = d.subplot2grid((3, 1), (2, 0))
d.setas = "x..y"
d.plot(fmt="g+")
d.title = ""
ax = d.subplot2grid((3, 1), (0, 0), rowspan=2)
d.setas = "xyy"
d.plot(fmt=["r.", "b-"])
d.xticklabels = [[]]
d.xlabel = ""
# Annotate plot with fitting parameters
d.annotate_fit(func, prefix="Model", x=0.7, y=0.3, fontdict={"size": "x-small"})
text = r"$y=A+Be^{-x/C}$" + "\n\n"
d.text(7.2, 3.9, text, fontdict={"size": "x-small"})
d.title = u"Differential Evolution Fit"
d["2nd-order polyfit coefficients"]),
fontdict={
"size": "x-small",
"color": "magenta"
},
)
d.setas = "xy"
d.curve_fit(SF.quadratic, result=True, header="Curve-fit")
d.setas = "x...y"
d.plot(fmt="b-", label="curve-fit")
d.annotate_fit(
SF.quadratic,
prefix="quadratic",
x=0.2,
y=0.65,
fontdict={
"size": "x-small",
"color": "blue"
},
)
d.setas = "xy"
fit = SF.Quadratic()
p0 = fit.guess(y, x=x)
d.lmfit(SF.Quadratic, p0=p0, result=True, header="lmfit")
d.setas = "x...y"
d.plot(fmt="g-", label="lmfit")
d.annotate_fit(
SF.Quadratic,
prefix="Quadratic",
"""Example of Arrhenius Fit.""" from Stoner import Data import Stoner.Fit as SF from numpy import linspace from numpy.random import normal #Make some data T=linspace(200,350,101) R=SF.arrhenius(T+normal(size=len(T),scale=1.0,loc=1.0),1E6,0.5) d=Data(T,R,setas="xy",column_headers=["T","Rate"]) #Curve_fit on its own d.curve_fit(SF.arrhenius,p0=[1E6,0.5],result=True,header="curve_fit") d.setas="xyy" d.plot() d.annotate_fit(SF.arrhenius,x=200,y=0.04) # lmfit using lmfit guesses fit=SF.Arrhenius() p0=fit.guess(R,x=T) d.lmfit(fit,p0=p0,result=True,header="lmfit") d.setas="x..y" d.plot() d.annotate_fit(SF.Arrhenius,x=200,y=0.02,prefix="Arrhenius") d.title="Arrhenius Test Fit" d.ylabel="Rate" d.xlabel="Temperature (K)"
ODRModel,
result=True,
header="ODR-Fit",
residuals=True,
output="report",
prefix="ODRModel",
)
# Reset labels
d.labels = []
# Make nice two panel plot layout
ax = d.subplot2grid((3, 1), (2, 0))
d.setas = "x..y"
d.plot(fmt="g+", label="Fit residuals")
d.setas = "x....y"
d.plot(fmt="b+", label="ODRModel Residuals")
d.title = ""
ax = d.subplot2grid((3, 1), (0, 0), rowspan=2)
d.setas = "xyy.y"
d.plot(fmt=["ro", "g-", "b-"])
d.xticklabels = [[]]
d.ax_xlabel = ""
# Annotate plot with fitting parameters
d.annotate_fit(PowerLaw, x=0.1, y=0.25, fontdict={"size": "x-small"})
d.annotate_fit(
ODRModel, x=0.65, y=0.15, fontdict={"size": "x-small"}, prefix="ODRModel"
)
d.title = u"curve_fit with models"
y = lorentzian_diff(x, *params) + normal(size=len(x), scale=0.5)
dy = ones_like(x) * 5e-3
d = Data(x, y, dy, setas="xye", column_headers=["Time", "Signal", "dM"])
d.curve_fit(lorentzian_diff, p0=copy(params), result=True, header="curve_fit")
d.setas = "xye"
d.lmfit(Lorentzian_diff, result=True, header="lmfit", prefix="lmfit")
d.setas = "xyeyy"
d.plot(fmt=["r+", "b-", "g-"])
d.annotate_fit(
lorentzian_diff,
x=0.6,
y=0.2,
fontdict={"size": "x-small", "color": "blue"},
mode="eng",
)
d.annotate_fit(
Lorentzian_diff,
x=0.05,
y=0.2,
fontdict={"size": "x-small", "color": "green"},
prefix="lmfit",
mode="eng",
)
d.title = "Differential Lorentzian Fit"
params = [265, 65, 1.0, 5]
params2 = deepcopy(params)
G = blochGrueneisen(T, *params) + normal(size=len(T), scale=5e-5)
dG = ones_like(T) * 5e-5
d = Data(
T,
G,
dG,
setas="xye",
column_headers=["Temperature (K)", "Resistivity", "dR"],
)
d.curve_fit(blochGrueneisen, p0=params, result=True, header="curve_fit")
d.setas = "xy"
d.lmfit(BlochGrueneisen, result=True, header="lmfit")
d.setas = "xyeyy"
d.plot(fmt=["r.", "b-", "g-"])
d.annotate_fit(blochGrueneisen, x=0.65, y=0.35, fontdict={"size": "x-small"})
d.annotate_fit(
BlochGrueneisen,
x=0.65,
y=0.05,
fontdict={"size": "x-small"},
prefix="BlochGrueneisen",
)
d.title = "Bloch-Grueneisen Fit"
d.tight_layout()
p0 = fit.guess(G, x=B)
for p, v in zip(p0, params):
p0[p].set(v)
p0[p].max = v * 5
p0[p].min = 0
p0[p].vary = p != "T"
d.lmfit(fit, p0=p0, result=True, header="lmfit")
d.setas = "xyeyy"
d.plot(fmt=["r.", "b-", "g-"])
d.annotate_fit(func,
x=0.1,
y=0.5,
fontdict={
"size": "x-small",
"color": "blue"
},
mode="eng")
d.annotate_fit(
SF.Langevin,
x=0.1,
y=0.25,
fontdict={
"size": "x-small",
"color": "green"
},
prefix="Langevin",
mode="eng",
)
d.title = "langevin Fit"
header="ODR-Fit",
residuals=True,
output="report",
prefix="ODRModel",
)
# Reset labels
d.labels = []
# Make nice two panel plot layout
ax = d.subplot2grid((3, 1), (2, 0))
d.setas = "x..y"
d.plot(fmt="g+", label="Fit residuals")
d.setas = "x....y"
d.plot(fmt="b+", label="ODRModel Residuals")
d.title = ""
ax = d.subplot2grid((3, 1), (0, 0), rowspan=2)
d.setas = "xyy.y"
d.plot(fmt=["ro", "g-", "b-"])
d.xticklabels = [[]]
d.ax_xlabel = ""
# Annotate plot with fitting parameters
d.annotate_fit(PowerLaw, x=0.1, y=0.25, fontdict={"size": "x-small"})
d.annotate_fit(ODRModel,
x=0.65,
y=0.15,
fontdict={"size": "x-small"},
prefix="ODRModel")
d.title = u"curve_fit with models"
"""Example of nDimArrhenius Fit.""" from Stoner import Data import Stoner.Fit as SF from numpy import linspace from numpy.random import normal # Make some data T = linspace(200, 350, 101) R = SF.modArrhenius(T, 1e6, 0.5, 1.5) * normal(scale=0.00005, loc=1.0, size=len(T)) d = Data(T, R, setas="xy", column_headers=["T", "Rate"]) # Curve_fit on its own d.curve_fit(SF.modArrhenius, p0=[1e6, 0.5, 1.5], result=True, header="curve_fit") d.setas = "xyy" d.plot(fmt=["r.", "b-"]) d.annotate_fit(SF.modArrhenius, x=0.2, y=0.5) # lmfit using lmfit guesses fit = SF.ModArrhenius() p0 = [1e6, 0.5, 1.5] d.lmfit(fit, p0=p0, result=True, header="lmfit") d.setas = "x..y" d.plot() d.annotate_fit(SF.ModArrhenius, x=0.2, y=0.25, prefix="ModArrhenius") d.title = "Modified Arrhenius Test Fit" d.ylabel = "Rate" d.xlabel = "Temperature (K)"
import Stoner.Fit as SF
from numpy import linspace, ones_like
from numpy.random import normal
from copy import deepcopy
T = linspace(4.2, 300, 101)
params = [265, 65, 1.0, 5]
params2 = deepcopy(params)
G = SF.blochGrueneisen(T, *params) + normal(size=len(T), scale=5e-5)
dG = ones_like(T) * 5e-5
d = Data(T, G, dG, setas="xye", column_headers=["Temperature (K)", "Resistivity", "dR"])
d.curve_fit(SF.blochGrueneisen, p0=params, result=True, header="curve_fit")
d.setas = "xy"
d.lmfit(SF.BlochGrueneisen, p0=params2, result=True, header="lmfit")
d.setas = "xyeyy"
d.plot(fmt=["r.", "b-", "g-"])
d.annotate_fit(SF.blochGrueneisen, x=0.65, y=0.35, fontdict={"size": "x-small"})
d.annotate_fit(
SF.BlochGrueneisen,
x=0.65,
y=0.05,
fontdict={"size": "x-small"},
prefix="BlochGrueneisen",
)
d.title = "Bloch-Grueneisen Fit"
d.tight_layout()
