We Keep Coding

How to determine unknown parameters of a differential equation based on the best fit to a data set in Python?

I am trying to fit different differential equations to a given data set with python. For this reason, I use the scipy package, respectively the solve_ivp function.
This works fine for me, as long as I have a rough estimate of the parameters (b= 0.005) included in the differential equations, e.g:
import matplotlib.pyplot as plt
from scipy.integrate import solve_ivp
import numpy as np
def f(x, y, b):
dydx= [-b[0] * y[0]]
return dydx
xspan= np.linspace(1, 500, 25)
yinit= [5]
b= [0.005]
sol= solve_ivp(lambda x, y: f(x, y, b),
[xspan[0], xspan[-1]], yinit, t_eval= xspan)
print(sol)
print("\n")
print(sol.t)
print(sol.y)
plt.plot(sol.t, sol.y[0], "b--")
However, what I like to achieve is, that the parameter b (or more parameters) is/are determined "automatically" based on the best fit of the solved differential equation to a given data set (x and y). Is there a way this can be done, for example by combining this example with the curve_fit function of scipy and how would this look?
Thank you in advance!

Yes, what you think about should work, it should be easy to plug together. You want to call
popt, pcov = scipy.optimize.curve_fit(curve, xdata, ydata, p0=[b0])
b = popt[0]
where you now have to define a function curve(x,*p) that transforms any list of point into a list of values according to the only parameter b.
def curve(x,b):
res = solve_ivp(odefun, [1,500], [5], t_eval=x, args = [b])
return res.y[0]
Add optional arguments for error tolerances as necessary.
To make this more realistic, make also the initial point a parameter. Then it also becomes more obvious where a list is expected and where single arguments. To get a proper fitting task add some random noise to the test data. Also make the fall to zero not so fast, so that the final plot still looks somewhat interesting.
from scipy.integrate import solve_ivp
from scipy.optimize import curve_fit
xmin,xmax = 1,500
def f(t, y, b):
dydt= -b * y
return dydt
def curve(t, b, y0):
sol= solve_ivp(lambda t, y: f(t, y, b),
[xmin, xmax], [y0], t_eval= t)
return sol.y[0]
xdata = np.linspace(xmin, xmax, 25)
ydata = np.exp(-0.02*xdata)+0.02*np.random.randn(*xdata.shape)
y0 = 5
b= 0.005
p0 = [b,y0]
popt, pcov = curve_fit(curve, xdata, ydata, p0=p0)
b, y0 = popt
print(f"b={b}, y0 = {y0}")
This returns
b=0.019975693539459473, y0 = 0.9757709108115179
Now plot the test data against the fitted curve

Curve fitting of Monod growth/degradation equations to the experimental data

So the problem that is being faced here is the curve fitting of the Monod equations to the experimental data. The model of bacteria growth and degradation of the organic carbon looks like this:
dX/dt = (u * S * X )/(K + S)
dS/dt = ((-1/Y) * u * S * X )/(K + S)
These equations are solved using the scipy odeint function. Results after integration are stored into two vectors, one for growth, and the another one for degradation. The next step is to curve fit this model to the experimentally observed data and estimate the model parameters: u, K and Y. Once the code is run, the following error is produced:
File "C:\ProgramData\Anaconda3\lib\site-packages\scipy\optimize\minpack.py", line 392, in leastsq
raise TypeError('Improper input: N=%s must not exceed M=%s' % (n, m))
TypeError: Improper input: N=3 must not exceed M=2"
For the convenience, curve fitting part is commented out, so the plot of the expected result can be generated. Bellow is the code sample:
import numpy as np
import matplotlib.pyplot as plt
from scipy.integrate import odeint
from scipy.optimize import curve_fit
"""Experimental data!"""
t_exp = np.array([0, 8, 24, 32, 48, 96, 168])
S_exp = np.array([5.5, 4.7, 3.7, 2.5, 1.5, 0.7, 0.5])
X_exp = np.array([10000, 17000, 30000, 40000, 60000, 76000, 80000])
"Model of the microbial growth and the TOC degradation"
# SETTING UP THE MODEL
def f(t, u, K, Y):
'Function that returns mutually dependent variables X and S'
def growth(x, t):
X = x[0]
S = x[1]
"Now differential equations are defined!"
dXdt = (u * S * X )/(K + S)
dSdt = ((-1/Y) * u * S * X )/(K + S)
return [dXdt, dSdt]
# INTEGRATING THE DIFFERENTIAL EQUATIONS
"initial Conditions"
init = [10000, 5]
results = odeint(growth, init, t)
"Taking out desired column vectors from results array"
return results[:,0], results[:,1]
# CURVE FITTING AND PARAMETER ESTIMATION
"""k, kcov = curve_fit(f, t_exp, [X_exp, S_exp], p0=(1, 2, 2))
u = k[0]
K = k[1]
Y = k[2]"""
# RESULTS OF THE MODEL WITH THE ESTIMATED MODEL PARAMETERS
t_mod = np.linspace(0, 168, 100)
compute = f(t_mod, 0.8, 75, 13700)# these fit quite well, but estimated manually
X_mod = compute[0]
S_mod = compute[1]
# PLOT OF THE MODEL AND THE OBSERVED DATA
fig = plt.figure()
ax1 = fig.add_subplot(111)
ax1.plot(t_exp, X_exp, "yo")
ax1.plot(t_mod, X_mod, "g--", linewidth=3)
ax1.set_ylabel("X")
ax2 = ax1.twinx()
ax2.plot(t_exp, S_exp, "mo", )
ax2.plot(t_mod, S_mod, "r--", linewidth=3)
ax2.set_ylabel("S", color="r")
for tl in ax2.get_yticklabels():
tl.set_color("r")
plt.show()
Any advice of how to cope with this problem and proceed further would be highly appreciated. Thanks in advance.

The result of f() needs to have the same shape as the experimental data you feed into curve_fit as third parameter. In the last line of f() you just take the t = 0s value of the solution for both ODEs and return that, but you should return the complete solution. When fitting several sets of data at once using curve_fit, just concat them (stack horizontally), i.e.
def f(t, u, K, Y):
.....
return np.hstack((results[:,0], results[:,1]))
and call curve_fit like
k, kcov = curve_fit(f, t_exp, np.hstack([X_exp, S_exp]), p0=(1, 2, 2))
You will have to adapt the plotting part of your script, too:
compute = f(t_mod, u, K, Y)
compute = compute.reshape((2,-1))

Python curve_fit with multiple independent variables

Python's curve_fit calculates the best-fit parameters for a function with a single independent variable, but is there a way, using curve_fit or something else, to fit for a function with multiple independent variables? For example:
def func(x, y, a, b, c):
return log(a) + b*log(x) + c*log(y)
where x and y are the independent variable and we would like to fit for a, b, and c.

You can pass curve_fit a multi-dimensional array for the independent variables, but then your func must accept the same thing. For example, calling this array X and unpacking it to x, y for clarity:
import numpy as np
from scipy.optimize import curve_fit
def func(X, a, b, c):
x,y = X
return np.log(a) + b*np.log(x) + c*np.log(y)
# some artificially noisy data to fit
x = np.linspace(0.1,1.1,101)
y = np.linspace(1.,2., 101)
a, b, c = 10., 4., 6.
z = func((x,y), a, b, c) * 1 + np.random.random(101) / 100
# initial guesses for a,b,c:
p0 = 8., 2., 7.
print(curve_fit(func, (x,y), z, p0))
Gives the fit:
(array([ 9.99933937, 3.99710083, 6.00875164]), array([[ 1.75295644e-03, 9.34724308e-05, -2.90150983e-04],
[ 9.34724308e-05, 5.09079478e-06, -1.53939905e-05],
[ -2.90150983e-04, -1.53939905e-05, 4.84935731e-05]]))

optimizing a function with multiple input dimensions and a variable number of parameters
This example shows how to fit a polynomial with a two dimensional input (R^2 -> R) by an increasing number of coefficients. The design is very flexible so that the callable f from curve_fit is defined once for any number of non-keyword arguments.
minimal reproducible example
import numpy as np
import matplotlib.pyplot as plt
from scipy.optimize import curve_fit
def poly2d(xy, *coefficients):
x = xy[:, 0]
y = xy[:, 1]
proj = x + y
res = 0
for order, coef in enumerate(coefficients):
res += coef * proj ** order
return res
nx = 31
ny = 21
range_x = [-1.5, 1.5]
range_y = [-1, 1]
target_coefficients = (3, 0, -19, 7)
xs = np.linspace(*range_x, nx)
ys = np.linspace(*range_y, ny)
im_x, im_y = np.meshgrid(xs, ys)
xdata = np.c_[im_x.flatten(), im_y.flatten()]
im_target = poly2d(xdata, *target_coefficients).reshape(ny, nx)
fig, axs = plt.subplots(2, 3, figsize=(29.7, 21))
axs = axs.flatten()
ax = axs[0]
ax.set_title('Unknown polynomial P(x+y)\n[secret coefficients: ' + str(target_coefficients) + ']')
sm = ax.imshow(
im_target,
cmap = plt.get_cmap('coolwarm'),
origin='lower'
)
fig.colorbar(sm, ax=ax)
for order in range(5):
ydata=im_target.flatten()
popt, pcov = curve_fit(poly2d, xdata=xdata, ydata=ydata, p0=[0]*(order+1) )
im_fit = poly2d(xdata, *popt).reshape(ny, nx)
ax = axs[1+order]
title = 'Fit O({:d}):'.format(order)
for o, p in enumerate(popt):
if o%2 == 0:
title += '\n'
if o == 0:
title += ' {:=-{w}.1f} (x+y)^{:d}'.format(p, o, w=int(np.log10(max(abs(p), 1))) + 5)
else:
title += ' {:=+{w}.1f} (x+y)^{:d}'.format(p, o, w=int(np.log10(max(abs(p), 1))) + 5)
title += '\nrms: {:.1f}'.format( np.mean((im_fit-im_target)**2)**.5 )
ax.set_title(title)
sm = ax.imshow(
im_fit,
cmap = plt.get_cmap('coolwarm'),
origin='lower'
)
fig.colorbar(sm, ax=ax)
for ax in axs.flatten():
ax.set_xlabel('x')
ax.set_ylabel('y')
plt.show()
P.S. The concept of this answer is identical to my other answer here, but the code example is way more clear. At the time given, I will delete the other answer.

Fitting to an unknown numer of parameters
In this example, we try to reproduce some measured data measData.
In this example measData is generated by the function measuredData(x, a=.2, b=-2, c=-.8, d=.1). I practice, we might have measured measData in a way - so we have no idea, how it is described mathematically. Hence the fit.
We fit by a polynomial, which is described by the function polynomFit(inp, *args). As we want to try out different orders of polynomials, it is important to be flexible in the number of input parameters.
The independent variables (x and y in your case) are encoded in the 'columns'/second dimension of inp.
import numpy as np
import matplotlib
import matplotlib.pyplot as plt
from scipy.optimize import curve_fit
def measuredData(inp, a=.2, b=-2, c=-.8, d=.1):
x=inp[:,0]
y=inp[:,1]
return a+b*x+c*x**2+d*x**3 +y
def polynomFit(inp, *args):
x=inp[:,0]
y=inp[:,1]
res=0
for order in range(len(args)):
print(14,order,args[order],x)
res+=args[order] * x**order
return res +y
inpData=np.linspace(0,10,20).reshape(-1,2)
inpDataStr=['({:.1f},{:.1f})'.format(a,b) for a,b in inpData]
measData=measuredData(inpData)
fig, ax = plt.subplots()
ax.plot(np.arange(inpData.shape[0]), measData, label='measuered', marker='o', linestyle='none' )
for order in range(5):
print(27,inpData)
print(28,measData)
popt, pcov = curve_fit(polynomFit, xdata=inpData, ydata=measData, p0=[0]*(order+1) )
fitData=polynomFit(inpData,*popt)
ax.plot(np.arange(inpData.shape[0]), fitData, label='polyn. fit, order '+str(order), linestyle='--' )
ax.legend( loc='upper left', bbox_to_anchor=(1.05, 1))
print(order, popt)
ax.set_xticklabels(inpDataStr, rotation=90)
Result:

Yes. We can pass multiple variables for curve_fit. I have written a piece of code:
import numpy as np
x = np.random.randn(2,100)
w = np.array([1.5,0.5]).reshape(1,2)
esp = np.random.randn(1,100)
y = np.dot(w,x)+esp
y = y.reshape(100,)
In the above code I have generated x a 2D data set in shape of (2,100) i.e, there are two variables with 100 data points. I have fit the dependent variable y with independent variables x with some noise.
def model_func(x,w1,w2,b):
w = np.array([w1,w2]).reshape(1,2)
b = np.array([b]).reshape(1,1)
y_p = np.dot(w,x)+b
return y_p.reshape(100,)
We have defined a model function that establishes relation between y & x.
Note: The shape of output of the model function or predicted y should be (length of x,)
popt, pcov = curve_fit(model_func,x,y)
The popt is an 1D numpy array containing predicted parameters. In our case there are 3 parameters.

Yes, there is: simply give curve_fit a multi-dimensional array for xData.

Using scipy.optimize.curve_fit with weights

According to the documentation, the argument sigma can be used to set the weights of the data points in the fit. These "describe" 1-sigma errors when the argument absolute_sigma=True.
I have some data with artificial normally-distributed noise which varies:
n = 200
x = np.linspace(1, 20, n)
x0, A, alpha = 12, 3, 3
def f(x, x0, A, alpha):
return A * np.exp(-((x-x0)/alpha)**2)
noise_sigma = x/20
noise = np.random.randn(n) * noise_sigma
yexact = f(x, x0, A, alpha)
y = yexact + noise
If I want to fit the noisy y to f using curve_fit to what should I set sigma? The documentation isn't very specific here, but I would usually use 1/noise_sigma**2 as the weight:
p0 = 10, 4, 2
popt, pcov = curve_fit(f, x, y, p0)
popt2, pcov2 = curve_fit(f, x, y, p0, sigma=1/noise_sigma**2, absolute_sigma=True)
It doesn't seem to improve the fit much, though.
Is this option only used to better interpret the fit uncertainties through the covariance matrix? What is the difference between these two telling me?
In [249]: pcov
Out[249]:
array([[ 1.10205238e-02, -3.91494024e-08, 8.81822412e-08],
[ -3.91494024e-08, 1.52660426e-02, -1.05907265e-02],
[ 8.81822412e-08, -1.05907265e-02, 2.20414887e-02]])
In [250]: pcov2
Out[250]:
array([[ 0.26584674, -0.01836064, -0.17867193],
[-0.01836064, 0.27833 , -0.1459469 ],
[-0.17867193, -0.1459469 , 0.38659059]])

At least with scipy version 1.1.0 the parameter sigma should be equal to the error on each parameter. Specifically the documentation says:
A 1-d sigma should contain values of standard deviations of errors in
ydata. In this case, the optimized function is chisq = sum((r / sigma)
** 2).
In your case that would be:
curve_fit(f, x, y, p0, sigma=noise_sigma, absolute_sigma=True)
I looked through the source code and verified that when you specify sigma this way it minimizes ((f-data)/sigma)**2.
As a side note, this is in general what you want to be minimizing when you know the errors. The likelihood of observing points data given a model f is given by:
L(data|x0,A,alpha) = product over i Gaus(data_i, mean=f(x_i,x0,A,alpha), sigma=sigma_i)
which if you take the negative log becomes (up to constant factors that don't depend on the parameters):
-log(L) = sum over i (f(x_i,x0,A,alpha)-data_i)**2/(sigma_i**2)
which is just the chisquare.
I wrote a test program to verify that curve_fit was indeed returning the correct values with the sigma specified correctly:
from __future__ import print_function
import numpy as np
from scipy.optimize import curve_fit, fmin
np.random.seed(0)
def make_chi2(x, data, sigma):
def chi2(args):
x0, A, alpha = args
return np.sum(((f(x,x0,A,alpha)-data)/sigma)**2)
return chi2
n = 200
x = np.linspace(1, 20, n)
x0, A, alpha = 12, 3, 3
def f(x, x0, A, alpha):
return A * np.exp(-((x-x0)/alpha)**2)
noise_sigma = x/20
noise = np.random.randn(n) * noise_sigma
yexact = f(x, x0, A, alpha)
y = yexact + noise
p0 = 10, 4, 2
# curve_fit without parameters (sigma is implicitly equal to one)
popt, pcov = curve_fit(f, x, y, p0)
# curve_fit with wrong sigma specified
popt2, pcov2 = curve_fit(f, x, y, p0, sigma=1/noise_sigma**2, absolute_sigma=True)
# curve_fit with correct sigma
popt3, pcov3 = curve_fit(f, x, y, p0, sigma=noise_sigma, absolute_sigma=True)
chi2 = make_chi2(x,y,noise_sigma)
# double checking that we get the correct answer
xopt = fmin(chi2,p0,xtol=1e-10,ftol=1e-10)
print("popt = %s, chi2 = %.2f" % (popt,chi2(popt)))
print("popt2 = %s, chi2 = %.2f" % (popt2, chi2(popt2)))
print("popt3 = %s, chi2 = %.2f" % (popt3, chi2(popt3)))
print("xopt = %s, chi2 = %.2f" % (xopt, chi2(xopt)))
which outputs:
popt = [ 11.93617403 3.30528488 2.86314641], chi2 = 200.66
popt2 = [ 11.94169083 3.30372955 2.86207253], chi2 = 200.64
popt3 = [ 11.93128545 3.333727 2.81403324], chi2 = 200.44
xopt = [ 11.93128603 3.33373094 2.81402741], chi2 = 200.44
As you can see the chi2 is indeed minimized correctly when you specify sigma=sigma as an argument to curve_fit.
As to why the improvement isn't "better", I'm not really sure. My only guess is that without specifying a sigma value you implicitly assume they are equal and over the part of the data where the fit matters (the peak), the errors are "approximately" equal.
To answer your second question, no the sigma option is not only used to change the output of the covariance matrix, it actually changes what is being minimized.

Python - calculating trendlines with errors

So I've got some data stored as two lists, and plotted them using
plot(datasetx, datasety)
Then I set a trendline
trend = polyfit(datasetx, datasety)
trendx = []
trendy = []
for a in range(datasetx[0], (datasetx[-1]+1)):
trendx.append(a)
trendy.append(trend[0]*a**2 + trend[1]*a + trend[2])
plot(trendx, trendy)
But I have a third list of data, which is the error in the original datasety. I'm fine with plotting the errorbars, but what I don't know is using this, how to find the error in the coefficients of the polynomial trendline.
So say my trendline came out to be 5x^2 + 3x + 4 = y, there needs to be some sort of error on the 5, 3 and 4 values.
Is there a tool using NumPy that will calculate this for me?

I think you can use the function curve_fit of scipy.optimize (documentation). A basic example of the usage:
import numpy as np
from scipy.optimize import curve_fit
def func(x, a, b, c):
return a*x**2 + b*x + c
x = np.linspace(0,4,50)
y = func(x, 5, 3, 4)
yn = y + 0.2*np.random.normal(size=len(x))
popt, pcov = curve_fit(func, x, yn)
Following the documentation, pcov gives:
The estimated covariance of popt. The diagonals provide the variance
of the parameter estimate.
So in this way you can calculate an error estimate on the coefficients. To have the standard deviation you can take the square root of the variance.
Now you have an error on the coefficients, but it is only based on the deviation between the ydata and the fit. In case you also want to account for an error on the ydata itself, the curve_fit function provides the sigma argument:
sigma : None or N-length sequence
If not None, it represents the standard-deviation of ydata. This
vector, if given, will be used as weights in the least-squares
problem.
A complete example:
import numpy as np
from scipy.optimize import curve_fit
def func(x, a, b, c):
return a*x**2 + b*x + c
x = np.linspace(0,4,20)
y = func(x, 5, 3, 4)
# generate noisy ydata
yn = y + 0.2 * y * np.random.normal(size=len(x))
# generate error on ydata
y_sigma = 0.2 * y * np.random.normal(size=len(x))
popt, pcov = curve_fit(func, x, yn, sigma = y_sigma)
# plot
import matplotlib.pyplot as plt
fig = plt.figure()
ax = fig.add_subplot(111)
ax.errorbar(x, yn, yerr = y_sigma, fmt = 'o')
ax.plot(x, np.polyval(popt, x), '-')
ax.text(0.5, 100, r"a = {0:.3f} +/- {1:.3f}".format(popt[0], pcov[0,0]**0.5))
ax.text(0.5, 90, r"b = {0:.3f} +/- {1:.3f}".format(popt[1], pcov[1,1]**0.5))
ax.text(0.5, 80, r"c = {0:.3f} +/- {1:.3f}".format(popt[2], pcov[2,2]**0.5))
ax.grid()
plt.show()
Then something else, about using numpy arrays. One of the main advantages of using numpy is that you can avoid for loops because operations on arrays apply elementwise. So the for-loop in your example can also be done as following:
trendx = arange(datasetx[0], (datasetx[-1]+1))
trendy = trend[0]*trendx**2 + trend[1]*trendx + trend[2]
Where I use arange instead of range as it returns a numpy array instead of a list.
In this case you can also use the numpy function polyval:
trendy = polyval(trend, trendx)

I have not been able to find any way of getting the errors in the coefficients that is built in to numpy or python. I have a simple tool that I wrote based on Section 8.5 and 8.6 of John Taylor's An Introduction to Error Analysis. Maybe this will be sufficient for your task (note the default return is the variance, not the standard deviation). You can get large errors (as in the provided example) because of significant covariance.
def leastSquares(xMat, yMat):
'''
Purpose
-------
Perform least squares using the procedure outlined in 8.5 and 8.6 of Taylor, solving
matrix equation X a = Y
Examples
--------
>>> from scipy import matrix
>>> xMat = matrix([[ 1, 5, 25],
[ 1, 7, 49],
[ 1, 9, 81],
[ 1, 11, 121]])
>>> # matrix has rows of format [constant, x, x^2]
>>> yMat = matrix([[142],
[168],
[211],
[251]])
>>> a, varCoef, yRes = leastSquares(xMat, yMat)
>>> # a is a column matrix, holding the three coefficients a, b, c, corresponding to
>>> # the equation a + b*x + c*x^2
Returns
-------
a: matrix
best fit coefficients
varCoef: matrix
variance of derived coefficents
yRes: matrix
y-residuals of fit
'''
xMatSize = xMat.shape
numMeas = xMatSize[0]
numVars = xMatSize[1]
xxMat = xMat.T * xMat
xyMat = xMat.T * yMat
xxMatI = xxMat.I
aMat = xxMatI * xyMat
yAvgMat = xMat * aMat
yRes = yMat - yAvgMat
var = (yRes.T * yRes) / (numMeas - numVars)
varCoef = xxMatI.diagonal() * var[0, 0]
return aMat, varCoef, yRes