fitting array parameter

[Zohreh Karimzadeh]

Hi lmfit

I stuck in my fitting due to my parameters,

Please kindly let me know how this issue could be solved.

----------------------------------------------------

from lmfit import Model
# (y, params, independent_variables=['x1, 'x2']assign initial value for parameters)

def cationshortrange(xn,wnca, neutral = ['water'], cation = ['H'], anion = ['I'], zc = [1], za = [1], ea = [0.1] ):
term_wnMa = 0
for k in range(len(anion)):
for j in range(len(cation)):
if j < 1:
for i in range(len(neutral)):
term_wnMa += xn[i] * ea[k] * ((zc[j] + za[k]) / (za[k])) * wnca[i, j, k]
return print(term_wnMa)
# cationshortrange(wnca[water, H,I] ==+ 0.5, xn = 0.1 , neutral = ['water'], cation = ['H', 'K'], anion = ['I', 'Br'], zc = [1, 1], za = [1, 1], ea = [0.1] )
xn = [1, 2, 3, 4, 5]
y = [0.1, 0.2, 1.2, 1.6, 2.0]
gmodel = Model(cationshortrange)
params = gmodel.make_params()
print('parameter names: {}'.format(gmodel.param_names))
print('independent variables: {}'.format(gmodel.independent_vars))
result = gmodel.fit(y, params)
print(result.fit_report())
print(result.params.items())

-------------------------------------------------------------

Traceback (most recent call last):
  File "C:\Users\Zohreh\PycharmProjects\PSCPureFit\dicParamLmfit.py", line 23, in <module>
    result = gmodel.fit(y, params)
  File "C:\Users\Zohreh\PycharmProjects\PSCPureFit\venv\lib\site-packages\lmfit\model.py", line 1023, in fit
    if not np.isscalar(kwargs[var]):
KeyError: 'xn'
parameter names: ['wnca']
independent variables: ['xn']

Process finished with exit code 1

----------------------------------------------------------------

In fact based on cation and anion I will have different number of parameters in wnca array.

[Matt Newville]

It looks to me like your model function `cationexchange` has one variable parameter: `wnca`.   All the other function arguments have default values that are lists and so would be treated as independent variables: those do not vary in the fit, and you will have to supply them to 

Model.fit().    It is possible that you intend some of `zc`, `za`, or `ea` to have values that vary in the fit, but that is not how lmfit work.  Each Parameter must have a unique name and a single floating-point (scalar) value.

Also, with

     params = gmodel.make_params()

or at some point before you use `params` in a fit,  you **must** provide initial values for all the variable parameters, in this case, `wnca`.  You cannot have initialized parameter values.

--
You received this message because you are subscribed to the Google Groups "lmfit-py" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lmfit-py+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lmfit-py/5be3db54-52fc-4858-ab5c-9f82310342f9n%40googlegroups.com.

--

--Matt Newville <newville at cars.uchicago.edu> 630-327-7411

[Zohreh Karimzadeh]

Thanks millions!

Zohreh Karimzadeh

https://www.researchgate.net/profile/Zohreh-Karimzadeh

Skype Name 49a52224a8b6b38b

Twitter Account @zohrehkarimzad1

z.kari...@gmail.com

+989102116325                                                        

((((((((((((((((Value Water)))))))))))))))

To view this discussion on the web visit https://groups.google.com/d/msgid/lmfit-py/CA%2B7ESbqE3MoQsqaEwxMAQt4wM-pj%3DEz--5vgLB4RH81j0X%3DBxw%40mail.gmail.com.