data-manipulation-utilities 0.2.8.dev956__tar.gz → 0.2.8.dev985__tar.gz

{data_manipulation_utilities-0.2.8.dev956 → data_manipulation_utilities-0.2.8.dev985}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: data-manipulation-utilities
-Version: 0.2.8.dev956
+Version: 0.2.8.dev985
 Summary: Project storing utilities needed to reduce boilerplate code when analyzing data
 Requires-Python: >=3.10,<3.13
 Classifier: Programming Language :: Python :: 3
@@ -1051,7 +1051,7 @@ utilities.placeholder_fit(kind='s+b', fit_dir='/some/directory')
 ### Likelihooods
 
 ```python
-from dmu.stats import utilities
+from dmu.stats import utilities as sut
 
 nll = sut.get_nll(kind='s+b')
 ```
@@ -1713,18 +1713,18 @@ plots:
         # Can add styling to specific plots, this should be the argument of
         # hist.plot(...)
         styling :
-        # This section will update the styling of each category
-        # The categories (class A, etc) are the keys of the dictionary of dataframes
-        class A:
-            # These are the arguments of plt.hist(...)
-            histtype : fill 
-            color    : gray
-            alpha    : 0.3
-        class B:
-            color    : red 
-            histtype : step
-            linestyle: '-'  # Linestyle is by default 'none', 
-                            # needs to be overriden to see _steps_
+            # This section will update the styling of each category
+            # The categories (class A, etc) are the keys of the dictionary of dataframes
+            class A:
+                # These are the arguments of plt.hist(...)
+                histtype : fill 
+                color    : gray
+                alpha    : 0.3
+            class B:
+                color    : red 
+                histtype : step
+                linestyle: '-'  # Linestyle is by default 'none', 
+                                # needs to be overriden to see _steps_
         # This will add vertical lines to plots, the arguments are the same
         # as the ones passed to axvline
         vline   :
@@ -1797,6 +1797,26 @@ plugin:
 Can be used to print statistics, mean, rms and weighted sum of entries for each distribution.
 The statistics, reffer to the data **only inside** the plotting range.
 
+#### Pulls
+
+If a given variable is a pull, one can add:
+
+- Fitted Gaussian
+- Fitter parameters
+- Lines representing the mean and width
+
+on the plot with:
+
+```yaml
+plugin:
+  pulls:
+    x_pul : {}
+```
+
+assuming that the variable holding the pulls is `x_pul`.
+The pulls will be drawn from -4 to +4. No configuration
+is available at the moment.
+
 ## 2D plots
 
 For the 2D case it would look like:

{data_manipulation_utilities-0.2.8.dev956 → data_manipulation_utilities-0.2.8.dev985}/README.md

@@ -1016,7 +1016,7 @@ utilities.placeholder_fit(kind='s+b', fit_dir='/some/directory')
 ### Likelihooods
 
 ```python
-from dmu.stats import utilities
+from dmu.stats import utilities as sut
 
 nll = sut.get_nll(kind='s+b')
 ```
@@ -1678,18 +1678,18 @@ plots:
         # Can add styling to specific plots, this should be the argument of
         # hist.plot(...)
         styling :
-        # This section will update the styling of each category
-        # The categories (class A, etc) are the keys of the dictionary of dataframes
-        class A:
-            # These are the arguments of plt.hist(...)
-            histtype : fill 
-            color    : gray
-            alpha    : 0.3
-        class B:
-            color    : red 
-            histtype : step
-            linestyle: '-'  # Linestyle is by default 'none', 
-                            # needs to be overriden to see _steps_
+            # This section will update the styling of each category
+            # The categories (class A, etc) are the keys of the dictionary of dataframes
+            class A:
+                # These are the arguments of plt.hist(...)
+                histtype : fill 
+                color    : gray
+                alpha    : 0.3
+            class B:
+                color    : red 
+                histtype : step
+                linestyle: '-'  # Linestyle is by default 'none', 
+                                # needs to be overriden to see _steps_
         # This will add vertical lines to plots, the arguments are the same
         # as the ones passed to axvline
         vline   :
@@ -1762,6 +1762,26 @@ plugin:
 Can be used to print statistics, mean, rms and weighted sum of entries for each distribution.
 The statistics, reffer to the data **only inside** the plotting range.
 
+#### Pulls
+
+If a given variable is a pull, one can add:
+
+- Fitted Gaussian
+- Fitter parameters
+- Lines representing the mean and width
+
+on the plot with:
+
+```yaml
+plugin:
+  pulls:
+    x_pul : {}
+```
+
+assuming that the variable holding the pulls is `x_pul`.
+The pulls will be drawn from -4 to +4. No configuration
+is available at the moment.
+
 ## 2D plots
 
 For the 2D case it would look like:

{data_manipulation_utilities-0.2.8.dev956 → data_manipulation_utilities-0.2.8.dev985}/pyproject.toml

@@ -1,7 +1,7 @@
 [tool.poetry]
 name           = 'data-manipulation-utilities'
 description    = 'Project storing utilities needed to reduce boilerplate code when analyzing data'
-version        = "0.2.8.dev956"
+version        = "0.2.8.dev985"
 readme         = 'README.md'
 package-mode   = true 
 packages       = [

{data_manipulation_utilities-0.2.8.dev956 → data_manipulation_utilities-0.2.8.dev985}/src/dmu/plotting/plotter_1d.py

@@ -9,6 +9,7 @@ from hist import Hist
 import numpy
 import matplotlib.pyplot as plt
 
+from scipy.stats           import norm
 from dmu.logging.log_store import LogStore
 from dmu.plotting.plotter  import Plotter
 from dmu.plotting.fwhm     import FWHM
@@ -59,6 +60,8 @@ class Plotter1D(Plotter):
 
         return minx, maxx, bins
     #-------------------------------------
+    # TODO: Might need to move all plugins code elsewhere
+    # and somehow register these plugins through config
     def _run_plugins(
             self,
             arr_val : numpy.ndarray,
@@ -66,7 +69,12 @@ class Plotter1D(Plotter):
             hst     : Hist,
             name    : str,
             varname : str) -> None:
-
+        '''
+        arr_val: Array of values of variable to plot
+        arr_wgt: Array of weights
+        hst    : Histogram, needed to plot extra information
+        name   : Name of variable
+        '''
         if 'plugin' not in self._d_cfg:
             log.debug('No plugins found')
             return
@@ -99,6 +107,56 @@ class Plotter1D(Plotter):
                 name    = name,
                 varname = varname,
                 cfg     = cfg)
+
+        if 'pulls' in self._d_cfg['plugin']:
+            if varname not in self._d_cfg['plugin']['pulls']:
+                log.debug(f'No pulls plugin found for variable {varname}')
+                return
+
+            log.debug(f'pulls plugin found for variable {varname}')
+            cfg = self._d_cfg['plugin']['pulls'][varname]
+            [minx, maxx, _] = self._d_cfg['plots' ][varname]['binning']
+
+            self._run_pulls(
+                arr_val = arr_val,
+                minx    = minx,
+                maxx    = maxx,
+                cfg     = cfg)
+    # ----------------------
+    def _run_pulls(
+        self,
+        arr_val : numpy.ndarray,
+        minx    : float,
+        maxx    : float,
+        cfg     : dict[str,str]) -> None:
+        '''
+        Parameters
+        -------------------
+        arr_val : Array of X axis values to plot
+        cfg     : Configuration for the statistics plugin
+        '''
+        mu, sg = norm.fit(arr_val)
+        nbins  = 200
+        arr_x  = numpy.linspace(minx, maxx, nbins)
+        arr_y  = norm.pdf(arr_x, mu, sg)
+        bw     = (maxx - minx) / nbins
+        area   = bw * arr_y.sum()
+        arr_y  = arr_y / area 
+
+        plt.plot(arr_x, arr_y, label='Fit', color='black')
+
+        stats = rf'''
+        $\mu={mu:.3f}$
+        $\sigma={sg:.3f}$
+        '''
+        ax = plt.gca()
+
+        maxy  = max(arr_y)
+        ax.text(x=-4.5, y=0.93 * maxy, s=stats, fontsize=30)
+
+        ax.axvline(x=mu     , ls=':', lw=1, c='r')
+        ax.axvline(x=mu - sg, ls='-', lw=1, c='r')
+        ax.axvline(x=mu + sg, ls='-', lw=1, c='r')
     # ----------------------
     def _trim_to_range(
         self,
@@ -278,7 +336,20 @@ class Plotter1D(Plotter):
 
         return f'{name:<15}{nentries:<10}'
     # --------------------------------------------
-    def _normalize_weights(self, arr_wgt : numpy.ndarray, var : str) -> numpy.ndarray:
+    def _normalize_weights(
+        self, 
+        arr_wgt : numpy.ndarray, 
+        var     : str) -> numpy.ndarray:
+        '''
+        Parameters
+        --------------
+        arr_wgt : Array of weights
+        var     : Plotting variable, needed to access normalization flag
+
+        Returns
+        --------------
+        Array of weights, normalized to 1 or not
+        '''
         cfg_var = self._d_cfg['plots'][var]
         if 'normalized' not in cfg_var:
             log.debug(f'Not normalizing for variable: {var}')
@@ -288,9 +359,12 @@ class Plotter1D(Plotter):
             log.debug(f'Not normalizing for variable: {var}')
             return arr_wgt
 
+        [minx, maxx, nbins] = self._d_cfg['plots'][var]['binning']
+
         log.debug(f'Normalizing for variable: {var}')
-        total   = numpy.sum(arr_wgt)
-        arr_wgt = arr_wgt / total
+        bw      = (maxx - minx) / nbins
+        area    = bw * numpy.sum(arr_wgt)
+        arr_wgt = arr_wgt / area 
 
         return arr_wgt
     # --------------------------------------------

{data_manipulation_utilities-0.2.8.dev956 → data_manipulation_utilities-0.2.8.dev985}/src/dmu/stats/fitter.py

@@ -374,7 +374,7 @@ class Fitter:
 
         stat = res.status
 
-        log.info(f'{chi2:<10.3f}{pval:<10.3e}{stat:<10}')
+        log.debug(f'{chi2:<10.3f}{pval:<10.3e}{stat:<10}')
         pdf   = nll.model[0] # This class is not meant for simultaneous fits
                              # There should only be one PDF
         d_par = Fitter.get_float_pars(pdf=pdf)

{data_manipulation_utilities-0.2.8.dev956 → data_manipulation_utilities-0.2.8.dev985}/src/dmu/stats/utilities.py

@@ -30,7 +30,8 @@ from zfit.interface   import ZfitSpace     as zobs
 from zfit.interface   import ZfitPDF       as zpdf
 from zfit.interface   import ZfitModel     as zmod
 from zfit.interface   import ZfitParameter as zpar
-from zfit.result      import FitResult     as zres
+
+from zfit.minimizers.interface   import ZfitResult    as zres
 
 log = LogStore.add_logger('dmu:stats:utilities')
 #-------------------------------------------------------
@@ -535,6 +536,7 @@ def pdf_to_tex(path : str, d_par : dict[str,str], skip_fixed : bool = True) -> N
 #---------------------------------------------
 def get_model(
     kind   : str,
+    nsample: int       = 1000,
     obs    : zobs|None = None,
     suffix : str|None  = None,
     lam    : float     = -0.0001) -> zpdf:
@@ -543,10 +545,11 @@ def get_model(
 
     Parameters:
 
-    kind  : 'signal' for Gaussian, 's+b' for Gaussian plus exponential
-    obs   : If provided, will use it, by default None and will be built in function
-    suffix: Optional, can be used in case multiple models are needed
-    lam   : Decay constant of exponential component, set to -0.0001 by default
+    kind   : 'signal' for Gaussian, 's+b' for Gaussian plus exponential
+    nsample: Number of entries for normalization of each component, default 1000
+    obs    : If provided, will use it, by default None and will be built in function
+    suffix : Optional, can be used in case multiple models are needed
+    lam    : Decay constant of exponential component, set to -0.0001 by default
     '''
     if suffix is not None:
         suffix = f'_{suffix}'
@@ -567,8 +570,8 @@ def get_model(
     expo= zfit.pdf.Exponential(obs=obs, lam=c)
 
     if kind == 's+b':
-        nexpo = zfit.param.Parameter(f'nbkg{suffix}', 1000, 0, 1000_000)
-        ngaus = zfit.param.Parameter(f'nsig{suffix}', 1000, 0, 1000_000)
+        nexpo = zfit.param.Parameter(f'nbkg{suffix}', nsample, 0, 1000_000)
+        ngaus = zfit.param.Parameter(f'nsig{suffix}', nsample, 0, 1000_000)
 
         bkg   = expo.create_extended(nexpo)
         sig   = gaus.create_extended(ngaus)
@@ -590,7 +593,6 @@ def get_nll(kind : str) -> ZfitLoss:
     '''
     pdf = get_model(kind=kind)
 
-
     if kind == 's+b':
         dat = pdf.create_sampler()
         return zfit.loss.ExtendedUnbinnedNLL(model=pdf, data=dat)

data_manipulation_utilities-0.2.8.dev985/src/dmu_data/plotting/tests/pulls.yaml

@@ -0,0 +1,18 @@
+saving:
+  plt_dir: plotting/pluggins/pulls
+plots:
+  x_pul:
+    binning   : [-4.0, 4.0, 60]
+    labels    : ['x Pull', 'Entries']
+    normalized: true
+    styling:
+      pull:
+        histtype : fill
+        color    : blue
+        alpha    : 0.3
+  y_pul:
+    binning: [-4.0, 4.0, 60]
+    labels : ['y Pull', 'Normalized']
+plugin:
+  pulls:
+    x_pul : {}