LabTools3 1.1.3.14__tar.gz → 1.1.3.16__tar.gz

{labtools3-1.1.3.14 → labtools3-1.1.3.16}/LT/box.py

@@ -456,8 +456,10 @@ class histo:
         Keyword        Meaning
         ============   =====================================================
         scale          True: the original bin number is not a multiple of n
-                             and the last bin content will be scaled
-        use_mean       True: the new bin content is the mean of the bin_content
+                             and the last bin content will be scaled accordingly
+                             
+        use_mean       True: the new bin content is the mean of the bin_content 
+                             of n bins
 
         replace        True: replace the current histogram with the
                              rebinned version
@@ -1138,7 +1140,7 @@ class histo:
         # mean value of the slices
         mean_sl = np.array([np.mean(x[sl]) for sl in slices[:-1]])
         # factor to correct the sum for slices that are shorter than n
-        fact = np.array([np.float(n)/len(x[sl]) for sl in slices[:-1]])
+        fact = np.array([float(n)/len(x[sl]) for sl in slices[:-1]])
         # return the values
         return sum_sl, mean_sl, slices[:-1], fact
 

{labtools3-1.1.3.14 → labtools3-1.1.3.16}/LT/parameterfile.py

@@ -39,6 +39,9 @@ Possible data types (dtype) are:
 If ``var_type`` keyword agrument is not used, the function tries to convert to float or bool. If both fails it
 returns the string value.
 
+An input line can be continued (will be concatenated) by adding a backslash as the last character or
+a comma. The comma will be included in the data while the backslash is removed.
+
 Variable names are also called keys.
 """
 
@@ -58,7 +61,7 @@ class pfile:
     >>> pf=pfile('my_datafile')
     
     """
-    def __init__(self, filename, data = None):
+    def __init__(self, filename = 'no_file_provided', data = None):
         """
         open a parameter file, scan for name=value pairs
         and return a dictionary with name value pairs
@@ -119,10 +122,13 @@ class pfile:
                 # check for continuation
                 if do_add :
                     # add new line to end of last one and reset do_add flag
-                    l_l = l_old[:-1] + l_l
+                    if l_old[-1] == ',' :  # preserve the comma as continuation character
+                        l_l = l_old[:] + l_l
+                    else:
+                        l_l = l_old[:-1] + l_l
                     do_add = False
-                # check for continuation character at the end of the line
-                do_cont = (l_l[-1] == '\\')
+                # check for continuation character at the end of the line (backslash or comma)
+                do_cont = (l_l[-1] == '\\' or l_l[-1] == ',')
                 if do_cont:
                    # found a continuation flag it and save the current line
                    do_add = True

{labtools3-1.1.3.14 → labtools3-1.1.3.16}/LT_Fit/gen_fit.py

@@ -10,23 +10,36 @@ Depending on the outcome, one might need to change the initial values of the par
 decide to change which parameters are fixed and which are to be fitted.
 
 1. Setting up the parameters::
-
+    
+    >>> import numpy as np
     >>> import LT_Fit.parameters as P  # get the parameter module
-    >>> import LT_fit.gen_fit as G     # load the genfit module
+    >>> import LT_Fit.gen_fit as G     # load the genfit module
     >>> a = P.Parameter(1., 'amp')     # parameter a, called 'amp', initialized to 1. 
     >>> b = P.Parameter(3., 'omega')   # parameter b, called 'omega', intialized to 3.
     >>> c = P.Parameter(1.5, 'phase')  # parameter c, called 'phase', initialize to 1.5
 
+
+    
+
 2. Define the fitting function::
 
     >>> def f(x):
-    ....    return a()*sin(b() * x + c() )
+    ....    return a()*np.sin(b() * x + c() )
+    
+
+3. Setup pseudo data::
+    
+    >>> xval = np.linspace(0,2,20)        # generate pseudo data
+    >>> y = f(xval)
+    >>> y_exp = y + np.random.normal(loc = 0, scale = 0.05, size = len(y))
+    >>> sig_y = np.ones_like(y)*0.05
+    
 
-3. Now do the fit::
+4. Now do the fit::
 
     >>> F = G.genfit( f, [a, b, c], x = xval, y = y_exp) #  the list [a,b,c] defines which parameters are fitted
-    >>> F = G.genfit( f, [a], x = xval, y = y_exp, yerr = sigy) #  only a is fitted, but take the exp. error into account
-    >>> F = G.genfit( f, [a, c], x = xval, y = y_exp, yerr = sigy) #  a and c are fitted b is kept constant
+    >>> F = G.genfit( f, [a], x = xval, y = y_exp, y_err = sig_y) #  only a is fitted, but take the exp. error into account
+    >>> F = G.genfit( f, [a, c], x = xval, y = y_exp, y_err = sig_y) #  a and c are fitted b is kept constant
     
 To change the initial values and repeat the fit::
 
@@ -34,11 +47,21 @@ To change the initial values and repeat the fit::
     >>> b.set(3.2)
     >>> c.set(2.0)
     >>> F = G.genfit( f, [a, b, c], x = xval, y = y_exp)
+    
+Calculate parameter derivatives analytically::
+    
+    >>> dfda = lambda x : np.sin(b() * x + c() )       # derivative wrsp a
+    >>> dfdb = lambda x : a()*np.cos(b() * x + c() )*x # derivative wrsp b
+    >>> dfdc = lambda x : a()*np.cos(b() * x + c() )   # derivative wrsp c
+    >>> deriv_list = [dfda, dfdb, dfdc]                # list of derivative functions
+    >>> # do the fit using analytically calculated derivatives
+    >>> F = G.genfit( f, [a, b, c], x = x, y = y_exp, y_err = sig_y, func_deriv=deriv_list) 
+
 
 Finally plot the data and the fit::
 
     >>> import LT.box as B
-    >>> B.plot_exp(xval, y_exp, sigy)   # plot the data with errorbars
+    >>> B.plot_exp(xval, y_exp, sig_y)   # plot the data with errorbars
     >>> F.plot()                        # plot the fitted function as a line    
     >>> B.plot_line(F.xpl, F.ypl)       # old version of plotting the fitted function as a line
     >>> F(x)                            # evaluate the fitted function at x
@@ -56,9 +79,9 @@ import scipy.stats as SS
 import matplotlib.pyplot as pl
 
 import copy as C
-from .parameters import *
+# from LT_Fit.parameters import *
 import numpy as np
-import pdb
+# import pdb
 
 
 class  genfit:
@@ -76,11 +99,16 @@ class  genfit:
     y_err        (:func:`numpy.array`) array of errors
     nplot        number of points to be used for plotting the fit
     ftol         minimal change in chi square to determine if the fit has converged
+    diff_step    step size used to the numerical calculation of derivetives with respect to the
+                 fit parameters (default = 0.001), h = diff_step * par
     kwargs       additional keywords are passes to scipy.optimize.least_squares, examples below
     bounds       an array for upper and lower bounds for the parameter
     loss         alternative loss function to handle outliers e.g. 'huber'
     f_scale      outliers with a residual more the f_scale should be not affect the result
     plot_fit     (default True) plot the fitted function automatically
+    func_deriv   list of functions corresponding to the fit parameter list (parameters)
+                 that calculate the derivative of the fit function with respect to each 
+                 of the fit parameters. (default = None)
     ==========   ===================================================================
 
     Additional keyword arguments are passed on to :func:`scipy.optimize.leastsq`
@@ -93,10 +121,12 @@ class  genfit:
                  nplot = 100, \
                  full_output = 1, \
                  ftol = 0.001, \
+                 diff_step = 0.001, \
                  print_results = True, \
                  plot_fit = True, \
+                 func_deriv = None, \
                  **kwargs):
-        # print('gen_fit kwargs = ', kwargs)
+        print('gen_fit kwargs = ', kwargs)
         self.plot_fit = plot_fit
         self.print_results = print_results
         self.y = y
@@ -111,18 +141,19 @@ class  genfit:
         self.nplot = nplot
         self.parameters = parameters # the array stores the addresses of the parameter objects
         self.func = function
+        self.func_deriv = func_deriv
         # carry out the fit
         if y is None:
             print('No values to fit, use set_yval to set them before fitting !')
             return
         return self.fit(full_output = full_output, \
                  ftol = ftol, \
+                 diff_step = diff_step, \
                  **kwargs)
 
     def f(self, params):
-        # define minimzation function for least square        i = 0
-        i = 0
-        for p in self.parameters:
+        # define minimzation function for least square
+        for i,p in enumerate(self.parameters):
             # store the current values in the array p back into the parameter
             # classes to be used by the user defined function
             # check if there is only 1 parameter
@@ -131,7 +162,6 @@ class  genfit:
                 p.set(params[i])
             else:
                 p.set(params)
-            i += 1
         # now calculate the difference between data and fit function
         # and return it to the optimization routine
         if self.y_err is None:
@@ -140,6 +170,24 @@ class  genfit:
             return (self.y - self.func(self.x))/self.y_err
     # end of the minimization function
     
+    def f_jac(self, params):
+        # calculate the jacobian analytically for a list of derivative functions if available
+        # define minimzation function for least square        i = 0
+        for i,p in enumerate(self.parameters):
+            # store the current values in the array p back into the parameter
+            # classes to be used by the user defined function
+            # check if there is only 1 parameter
+            ps = params.shape
+            if len(ps) > 0:
+                p.set(params[i])
+            else:
+                p.set(params)
+        if self.y_err is None:        
+            jac = np.array([-f(self.x) for f in self.func_deriv]).T
+        else:
+            jac = np.array([-f(self.x)/self.y_err for f in self.func_deriv]).T
+        return jac
+    
     def plot(self, xv = None, **kwargs):
         """
         Plot the fitting function 
@@ -162,7 +210,7 @@ class  genfit:
         else:
             pl.plot(xv, self.func(xv), **kwargs)    
 
-    def fit(self, full_output = 1, ftol = 0.001,  **kwargs):
+    def fit(self, full_output = 1, ftol = 0.001, diff_step = 0.001, **kwargs):
         # this is the minimzation routine
         # store the current parameter values in a list to be passed to the fitting function 
         # this is an implicit loop, it is equivalent to
@@ -170,6 +218,7 @@ class  genfit:
         # for param in parameters:
         #     p.append(param())
         # clear the parameter errors:
+        print('gen_fit.fit kwargs = ', kwargs)
         for p in self.parameters:
             p.err = 0.
         p = [param() for param in self.parameters]
@@ -178,12 +227,35 @@ class  genfit:
 #                                            full_output = full_output, \
 #                                            ftol = ftol, \
 #                                            **kwargs)
-        self.fit_result = optimize.least_squares( self.f, p,\
-                                            ftol = ftol, \
-                                            **kwargs)
+        if self.func_deriv is None:
+            print(f'Calculate numerical parameter derivatives with diff_step = {diff_step}')
+            self.fit_result = optimize.least_squares( self.f, p,\
+                                                ftol = ftol, \
+                                                diff_step = diff_step, \
+                                                **kwargs)
+        else:
+            print('Calculate analytical parameter derivatives ')
+            self.fit_result = optimize.least_squares( self.f, p,\
+                                                ftol = ftol, \
+                                                jac = self.f_jac, \
+                                                **kwargs)
+            
+            
         # estimate covariance matrix
-        J = self.fit_result.jac     
-        self.covar = np.linalg.inv(J.T.dot(J))
+        J = self.fit_result.jac 
+        try:
+            self.covar = np.linalg.inv(J.T.dot(J))
+        except Exception as err:
+            print(f'Cannot calculate covariance matrix, reason: {err}')
+            print(f'jacobian = {J}')
+            self.stat = {}
+            self.covar = [] 
+            self.xpl = []
+            self.ypl = []
+            self.chi2 = -1.
+            self.chi2_red = -1.
+            self.CL = -1.
+            return    
         # now calculate the fitted values
         fit_func = self.func(self.x)
         # final total chi square

{labtools3-1.1.3.14 → labtools3-1.1.3.16}/LabTools3.egg-info/PKG-INFO

@@ -1,14 +1,14 @@
 Metadata-Version: 2.1
 Name: LabTools3
-Version: 1.1.3.14
+Version: 1.1.3.16
 Summary: Python 3 Package of modules for typical analysis tasks analyzing physics data
 Home-page: http://wanda.fiu.edu/LabTools3
 Author: Werner Boeglin
 Author-email: boeglinw@fiu.edu
 License: MIT
 Keywords: Data Analysis
-Classifier: Programming Language :: Python :: 3.6
-Requires-Python: >=3.6
+Classifier: Programming Language :: Python :: 3.8
+Requires-Python: >=3.8
 Description-Content-Type: text/markdown
 License-File: LICENSE
 
@@ -50,3 +50,7 @@ Version 1.1.3.12: bug fixes, peak and background functions are available for his
 Version 1.1.3.13: fitting of only one parameters enabled for linear fits
 
 Version 1.1.3.14: error calcualtion for filling histograms with weights include sum of weight**2 
+
+Version 1.1.3.15: contination lines enabled in parameter file by endine line with "\" or ","" 
+
+Version 1.1.3.16: Corrected bug in histo.rebin. gen_fit allows control of relative step size using the diff_step keyword. It is also possible to supply a list of functions to calculate the parameter derivatives. This helps with numerical accuracy in the minimization process (see gen_fit documentation) 

labtools3-1.1.3.14/README.md → labtools3-1.1.3.16/PKG-INFO

@@ -1,3 +1,17 @@
+Metadata-Version: 2.1
+Name: LabTools3
+Version: 1.1.3.16
+Summary: Python 3 Package of modules for typical analysis tasks analyzing physics data
+Home-page: http://wanda.fiu.edu/LabTools3
+Author: Werner Boeglin
+Author-email: boeglinw@fiu.edu
+License: MIT
+Keywords: Data Analysis
+Classifier: Programming Language :: Python :: 3.8
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+
 # LabTools3
 Set of Python analysis tools for physics labs. It contains the "package" directory which is the Python package with a setup.py script for installation and
 a "doc" directory containing the Sphinx documentation. The repository starts with version 0.2.9 of LabTools.
@@ -35,4 +49,8 @@ Version 1.1.3.12: bug fixes, peak and background functions are available for his
 
 Version 1.1.3.13: fitting of only one parameters enabled for linear fits
 
-Version 1.1.3.14: error calcualtion for filling histograms with weights include sum of weight**2 
+Version 1.1.3.14: error calcualtion for filling histograms with weights include sum of weight**2 
+
+Version 1.1.3.15: contination lines enabled in parameter file by endine line with "\" or ","" 
+
+Version 1.1.3.16: Corrected bug in histo.rebin. gen_fit allows control of relative step size using the diff_step keyword. It is also possible to supply a list of functions to calculate the parameter derivatives. This helps with numerical accuracy in the minimization process (see gen_fit documentation) 

labtools3-1.1.3.14/PKG-INFO → labtools3-1.1.3.16/README.md

@@ -1,17 +1,3 @@
-Metadata-Version: 2.1
-Name: LabTools3
-Version: 1.1.3.14
-Summary: Python 3 Package of modules for typical analysis tasks analyzing physics data
-Home-page: http://wanda.fiu.edu/LabTools3
-Author: Werner Boeglin
-Author-email: boeglinw@fiu.edu
-License: MIT
-Keywords: Data Analysis
-Classifier: Programming Language :: Python :: 3.6
-Requires-Python: >=3.6
-Description-Content-Type: text/markdown
-License-File: LICENSE
-
 # LabTools3
 Set of Python analysis tools for physics labs. It contains the "package" directory which is the Python package with a setup.py script for installation and
 a "doc" directory containing the Sphinx documentation. The repository starts with version 0.2.9 of LabTools.
@@ -50,3 +36,7 @@ Version 1.1.3.12: bug fixes, peak and background functions are available for his
 Version 1.1.3.13: fitting of only one parameters enabled for linear fits
 
 Version 1.1.3.14: error calcualtion for filling histograms with weights include sum of weight**2 
+
+Version 1.1.3.15: contination lines enabled in parameter file by endine line with "\" or ","" 
+
+Version 1.1.3.16: Corrected bug in histo.rebin. gen_fit allows control of relative step size using the diff_step keyword. It is also possible to supply a list of functions to calculate the parameter derivatives. This helps with numerical accuracy in the minimization process (see gen_fit documentation) 

{labtools3-1.1.3.14 → labtools3-1.1.3.16}/setup.py

@@ -5,7 +5,7 @@ with open("README.md", "r") as fh:
 
 setup(
     name = "LabTools3",
-    version = "1.1.3.14",
+    version = "1.1.3.16",
     packages = find_packages(),
     # add additional files
     package_data = {'':['*.bat','*.command']},
@@ -20,7 +20,7 @@ setup(
     keywords = "Data Analysis",
     url = "http://wanda.fiu.edu/LabTools3",   # project home page, if any
     classifiers=[
-        "Programming Language :: Python :: 3.6",
+        "Programming Language :: Python :: 3.8",
         ],
-    python_requires='>=3.6'
+    python_requires='>=3.8'
 )