Source code for sas.sasgui.perspectives.fitting.fitting

"""
    Fitting perspective
"""
################################################################################
#This software was developed by the University of Tennessee as part of the
#Distributed Data Analysis of Neutron Scattering Experiments (DANSE)
#project funded by the US National Science Foundation.
#
#See the license text in license.txt
#
#copyright 2009, University of Tennessee
################################################################################
from __future__ import print_function

import re
import sys
import os
import wx
import logging
import numpy as np
import time
from copy import deepcopy
import traceback

import bumps.options
from bumps.gui.fit_dialog import show_fit_config
try:
    from bumps.gui.fit_dialog import EVT_FITTER_CHANGED
except ImportError:
    # CRUFT: bumps 0.7.5.8 and below
    EVT_FITTER_CHANGED = None  # type: wx.PyCommandEvent

from sas.sascalc.dataloader.loader import Loader
from sas.sascalc.fit.BumpsFitting import BumpsFit as Fit
from sas.sascalc.fit.pagestate import Reader, PageState, SimFitPageState
from sas.sascalc.fit import models

from sas.sasgui.guiframe.dataFitting import Data2D
from sas.sasgui.guiframe.dataFitting import Data1D
from sas.sasgui.guiframe.dataFitting import check_data_validity
from sas.sasgui.guiframe.events import NewPlotEvent
from sas.sasgui.guiframe.events import StatusEvent
from sas.sasgui.guiframe.events import EVT_SLICER_PANEL
from sas.sasgui.guiframe.events import EVT_SLICER_PARS_UPDATE
from sas.sasgui.guiframe.gui_style import GUIFRAME_ID
from sas.sasgui.guiframe.plugin_base import PluginBase
from sas.sasgui.guiframe.data_processor import BatchCell
from sas.sasgui.guiframe.gui_manager import MDIFrame
from sas.sasgui.guiframe.documentation_window import DocumentationWindow

from sas.sasgui.perspectives.calculator.model_editor import TextDialog
from sas.sasgui.perspectives.calculator.model_editor import EditorWindow
from sas.sasgui.perspectives.calculator.pyconsole import PyConsole

from .fitting_widgets import DataDialog
from .fit_thread import FitThread
from .fitpage import Chi2UpdateEvent
from .console import ConsoleUpdate
from .fitproblem import FitProblemDictionary
from .fitpanel import FitPanel
from .model_thread import Calc1D, Calc2D
from .resultpanel import ResultPanel, PlotResultEvent
from .gpu_options import GpuOptions

logger = logging.getLogger(__name__)

MAX_NBR_DATA = 4

(PageInfoEvent, EVT_PAGE_INFO) = wx.lib.newevent.NewEvent()


if sys.platform == "win32":
    ON_MAC = False
else:
    ON_MAC = True


[docs]class Plugin(PluginBase): """ Fitting plugin is used to perform fit """ def __init__(self): PluginBase.__init__(self, name="Fitting") #list of panel to send to guiframe self.mypanels = [] # reference to the current running thread self.calc_2D = None self.calc_1D = None self.color_dict = {} self.fit_thread_list = {} self.residuals = None self.weight = None self.fit_panel = None self.plot_panel = None # Start with a good default self.elapsed = 0.022 self.fit_panel = None ## dictionary of page closed and id self.closed_page_dict = {} ## Relative error desired in the sum of squares (float) self.batch_reset_flag = True #List of selected data self.selected_data_list = [] ## list of slicer panel created to display slicer parameters and results self.slicer_panels = [] # model 2D view self.model2D_id = None #keep reference of the simultaneous fit page self.sim_page = None self.sim_menu = None self.batch_page = None self.batch_menu = None self.index_model = 0 self.test_model_color = None #Create a reader for fit page's state self.state_reader = None self._extensions = '.fitv' self.menu1 = None self.new_model_frame = None self.temp_state = [] self.state_index = 0 self.sfile_ext = None # take care of saving data, model and page associated with each other self.page_finder = {} # Log startup logger.info("Fitting plug-in started") self.batch_capable = self.get_batch_capable()
[docs] def get_batch_capable(self): """ Check if the plugin has a batch capability """ return True
[docs] def create_fit_problem(self, page_id): """ Given an ID create a fitproblem container """ self.page_finder[page_id] = FitProblemDictionary()
[docs] def delete_fit_problem(self, page_id): """ Given an ID create a fitproblem container """ if page_id in self.page_finder.iterkeys(): del self.page_finder[page_id]
[docs] def add_color(self, color, id): """ adds a color as a key with a plot id as its value to a dictionary """ self.color_dict[id] = color
[docs] def on_batch_selection(self, flag): """ switch the the notebook of batch mode or not """ self.batch_on = flag if self.fit_panel is not None: self.fit_panel.batch_on = self.batch_on
[docs] def populate_menu(self, owner): """ Create a menu for the Fitting plug-in :param id: id to create a menu :param owner: owner of menu :return: list of information to populate the main menu """ #Menu for fitting self.menu1 = wx.Menu() id1 = wx.NewId() simul_help = "Add new fit panel" self.menu1.Append(id1, '&New Fit Page', simul_help) wx.EVT_MENU(owner, id1, self.on_add_new_page) self.menu1.AppendSeparator() self.id_simfit = wx.NewId() simul_help = "Constrained or Simultaneous Fit" self.menu1.Append(self.id_simfit, '&Constrained or Simultaneous Fit', simul_help) wx.EVT_MENU(owner, self.id_simfit, self.on_add_sim_page) self.sim_menu = self.menu1.FindItemById(self.id_simfit) self.sim_menu.Enable(False) #combined Batch self.id_batchfit = wx.NewId() batch_help = "Combined Batch" self.menu1.Append(self.id_batchfit, '&Combine Batch Fit', batch_help) wx.EVT_MENU(owner, self.id_batchfit, self.on_add_sim_page) self.batch_menu = self.menu1.FindItemById(self.id_batchfit) self.batch_menu.Enable(False) self.menu1.AppendSeparator() self.id_bumps_options = wx.NewId() bopts_help = "Fitting options" self.menu1.Append(self.id_bumps_options, 'Fit &Options', bopts_help) wx.EVT_MENU(owner, self.id_bumps_options, self.on_bumps_options) self.bumps_options_menu = self.menu1.FindItemById(self.id_bumps_options) self.bumps_options_menu.Enable(True) self.id_gpu_options_panel = wx.NewId() self.menu1.Append(self.id_gpu_options_panel, "OpenCL Options", "Choose OpenCL driver or turn it off") wx.EVT_MENU(owner, self.id_gpu_options_panel, self.on_gpu_options) self.id_result_panel = wx.NewId() self.menu1.Append(self.id_result_panel, "Fit Results", "Show fit results panel") wx.EVT_MENU(owner, self.id_result_panel, self.on_fit_results) self.menu1.AppendSeparator() self.id_reset_flag = wx.NewId() resetf_help = "BatchFit: If checked, the initial param values will be " resetf_help += "propagated from the previous results. " resetf_help += "Otherwise, the same initial param values will be used " resetf_help += "for all fittings." self.menu1.AppendCheckItem(self.id_reset_flag, "Chain Fitting [BatchFit Only]", resetf_help) wx.EVT_MENU(owner, self.id_reset_flag, self.on_reset_batch_flag) chain_menu = self.menu1.FindItemById(self.id_reset_flag) chain_menu.Check(not self.batch_reset_flag) chain_menu.Enable(self.batch_on) self.menu1.AppendSeparator() self.edit_model_menu = wx.Menu() # Find and put files name in menu try: self.set_edit_menu(owner=owner) except: raise self.id_edit = wx.NewId() self.menu1.AppendMenu(self.id_edit, "Plugin Model Operations", self.edit_model_menu) #create menubar items return [(self.menu1, self.sub_menu)]
[docs] def edit_custom_model(self, event): """ Get the python editor panel """ event_id = event.GetId() label = self.edit_menu.GetLabel(event_id) filename = os.path.join(models.find_plugins_dir(), label) frame = PyConsole(parent=self.parent, manager=self, panel=self.fit_panel, title='Advanced Plugin Model Editor', filename=filename) self.put_icon(frame) frame.Show(True)
[docs] def delete_custom_model(self, event): """ Delete custom model file """ event_id = event.GetId() label = self.delete_menu.GetLabel(event_id) toks = os.path.splitext(label) path = os.path.join(models.find_plugins_dir(), toks[0]) message = "Are you sure you want to delete the file {}?".format(path) dlg = wx.MessageDialog(self.frame, message, '', wx.YES_NO | wx.ICON_QUESTION) if not dlg.ShowModal() == wx.ID_YES: return try: for ext in ['.py', '.pyc']: p_path = path + ext if ext == '.pyc' and not os.path.isfile(path + ext): # If model is invalid, .pyc file may not exist as model has # never been compiled. Don't try and delete it continue os.remove(p_path) self.update_custom_combo() if os.path.isfile(p_path): msg = "Sorry! unable to delete the default " msg += "plugin model... \n" msg += "Please manually remove the files (.py, .pyc) " msg += "in the 'plugin_models' folder \n" msg += "inside of the SasView application, " msg += "and try it again." wx.MessageBox(msg, 'Info') #evt = StatusEvent(status=msg, type='stop', info='warning') #wx.PostEvent(self.parent, evt) else: self.delete_menu.Delete(event_id) for item in self.edit_menu.GetMenuItems(): if item.GetLabel() == label: self.edit_menu.DeleteItem(item) msg = "The plugin model, %s, has been deleted." % label evt = StatusEvent(status=msg, type='stop', info='info') wx.PostEvent(self.parent, evt) break except Exception: traceback.print_exc() msg = 'Delete Error: \nCould not delete the file; Check if in use.' wx.MessageBox(msg, 'Error')
[docs] def make_sum_model(self, event): """ Edit summodel template and make one """ event_id = event.GetId() model_manager = models.ModelManager() model_list = model_manager.composable_models() plug_dir = models.find_plugins_dir() textdial = TextDialog(None, self, wx.ID_ANY, 'Easy Sum/Multi(p1, p2) Editor', model_list, plug_dir) self.put_icon(textdial) textdial.ShowModal() textdial.Destroy()
[docs] def make_new_model(self, event): """ Make new model """ if self.new_model_frame is not None: self.new_model_frame.Show(False) self.new_model_frame.Show(True) else: event_id = event.GetId() dir_path = models.find_plugins_dir() title = "New Plugin Model Function" self.new_model_frame = EditorWindow(parent=self, base=self, path=dir_path, title=title) self.put_icon(self.new_model_frame) self.new_model_frame.Show(True)
[docs] def load_plugin_models(self, event): """ Update of models in plugin_models folder """ event_id = event.GetId() self.update_custom_combo()
[docs] def update_custom_combo(self): """ Update custom model list in the fitpage combo box """ try: # Update edit menus self.set_edit_menu_helper(self.parent, self.edit_custom_model) self.set_edit_menu_helper(self.parent, self.delete_custom_model) new_pmodel_list = self.fit_panel.reset_pmodel_list() if not new_pmodel_list: return # Redraws to a page not in focus are showing up as if they are # in the current page tab. current_page_index = self.fit_panel.GetSelection() current_page = self.fit_panel.GetCurrentPage() last_drawn_page = current_page # Set the new plugin model list for all fit pages; anticipating # categories, the updated plugin may be in either the form factor # or the structure factor combo boxes for uid, page in self.fit_panel.opened_pages.iteritems(): pbox = getattr(page, "formfactorbox", None) sbox = getattr(page, "structurebox", None) if pbox is None: continue # Set the new model list for the page page.model_list_box = new_pmodel_list # Grab names of the P and S models from the page. Need to do # this before resetting the form factor box since that clears # the structure factor box. old_struct = old_form = None form_name = pbox.GetValue() struct_name = sbox.GetStringSelection() if form_name: old_form = pbox.GetClientData(pbox.GetCurrentSelection()) if struct_name: old_struct = sbox.GetClientData(sbox.GetCurrentSelection()) # Reset form factor combo box. We are doing this for all # categories not just plugins since eventually the category # manager will allow plugin models to be anywhere. page._show_combox_helper() form_index = pbox.FindString(form_name) pbox.SetSelection(form_index) new_form = (pbox.GetClientData(form_index) if form_index != wx.NOT_FOUND else None) #print("form: %r"%form_name, old_form, new_form) # Reset structure factor combo box; even if the model list # hasn't changed, the model may have. Show the structure # factor combobox if the selected model is a form factor. sbox.Clear() page.initialize_combox() if new_form is not None and getattr(new_form, 'is_form_factor', False): sbox.Show() sbox.Enable() page.text2.Show() page.text2.Enable() struct_index = sbox.FindString(struct_name) sbox.SetSelection(struct_index) new_struct = (sbox.GetClientData(struct_index) if struct_index != wx.NOT_FOUND else None) #print("struct: %r"%struct_name, old_struct, new_struct) # Update the page if P or S has changed if old_form != new_form or old_struct != new_struct: #print("triggering model update") page._on_select_model(keep_pars=True) last_drawn_page = page # If last drawn is not the current, then switch the current to the # last drawn then switch back. Very ugly. if last_drawn_page != current_page: for page_index in range(self.fit_panel.PageCount): if self.fit_panel.GetPage(page_index) == last_drawn_page: self.fit_panel.SetSelection(page_index) break self.fit_panel.SetSelection(current_page_index) except Exception: logger.error("update_custom_combo: %s", sys.exc_value)
[docs] def set_edit_menu(self, owner): """ Set list of the edit model menu labels """ wx_id = wx.NewId() #new_model_menu = wx.Menu() self.edit_model_menu.Append(wx_id, 'New Plugin Model', 'Add a new model function') wx.EVT_MENU(owner, wx_id, self.make_new_model) wx_id = wx.NewId() self.edit_model_menu.Append(wx_id, 'Sum|Multi(p1, p2)', 'Sum of two model functions') wx.EVT_MENU(owner, wx_id, self.make_sum_model) e_id = wx.NewId() self.edit_menu = wx.Menu() self.edit_model_menu.AppendMenu(e_id, 'Advanced Plugin Editor', self.edit_menu) self.set_edit_menu_helper(owner, self.edit_custom_model) d_id = wx.NewId() self.delete_menu = wx.Menu() self.edit_model_menu.AppendMenu(d_id, 'Delete Plugin Models', self.delete_menu) self.set_edit_menu_helper(owner, self.delete_custom_model) wx_id = wx.NewId() self.edit_model_menu.Append(wx_id, 'Load Plugin Models', '(Re)Load all models present in user plugin_models folder') wx.EVT_MENU(owner, wx_id, self.load_plugin_models)
[docs] def set_edit_menu_helper(self, owner=None, menu=None): """ help for setting list of the edit model menu labels """ if menu is None: menu = self.edit_custom_model list_fnames = os.listdir(models.find_plugins_dir()) list_fnames.sort() for f_item in list_fnames: name = os.path.basename(f_item) toks = os.path.splitext(name) if toks[-1] == '.py' and not toks[0] == '__init__': if menu == self.edit_custom_model: if toks[0] == 'easy_sum_of_p1_p2': continue submenu = self.edit_menu else: submenu = self.delete_menu has_file = False for item in submenu.GetMenuItems(): if name == submenu.GetLabel(item.GetId()): has_file = True if not has_file: wx_id = wx.NewId() submenu.Append(wx_id, name) wx.EVT_MENU(owner, wx_id, menu) has_file = False
[docs] def put_icon(self, frame): """ Put icon in the frame title bar """ if hasattr(frame, "IsIconized"): if not frame.IsIconized(): try: icon = self.parent.GetIcon() frame.SetIcon(icon) except: pass
[docs] def on_add_sim_page(self, event): """ Create a page to access simultaneous fit option """ event_id = event.GetId() caption = "Const & Simul Fit" page = self.sim_page if event_id == self.id_batchfit: caption = "Combined Batch" page = self.batch_page def set_focus_page(page): page.Show(True) page.Refresh() page.SetFocus() #self.parent._mgr.Update() msg = "%s already opened\n" % str(page.window_caption) wx.PostEvent(self.parent, StatusEvent(status=msg)) if page is not None: return set_focus_page(page) if caption == "Const & Simul Fit": self.sim_page = self.fit_panel.add_sim_page(caption=caption) else: self.batch_page = self.fit_panel.add_sim_page(caption=caption)
[docs] def get_context_menu(self, plotpanel=None): """ Get the context menu items available for P(r).them allow fitting option for Data2D and Data1D only. :param graph: the Graph object to which we attach the context menu :return: a list of menu items with call-back function :note: if Data1D was generated from Theory1D the fitting option is not allowed """ self.plot_panel = plotpanel graph = plotpanel.graph fit_option = "Select data for fitting" fit_hint = "Dialog with fitting parameters " if graph.selected_plottable not in plotpanel.plots: return [] item = plotpanel.plots[graph.selected_plottable] if item.__class__.__name__ is "Data2D": if hasattr(item, "is_data"): if item.is_data: return [[fit_option, fit_hint, self._onSelect]] else: return [] return [[fit_option, fit_hint, self._onSelect]] else: # if is_data is true , this in an actual data loaded #else it is a data created from a theory model if hasattr(item, "is_data"): if item.is_data: return [[fit_option, fit_hint, self._onSelect]] else: return [] return [[fit_option, fit_hint, self._onSelect]] return []
[docs] def get_panels(self, parent): """ Create and return a list of panel objects """ self.parent = parent #self.parent.Bind(EVT_FITSTATE_UPDATE, self.on_set_state_helper) # Creation of the fit panel self.frame = MDIFrame(self.parent, None, 'None', (100, 200)) self.fit_panel = FitPanel(parent=self.frame, manager=self) self.frame.set_panel(self.fit_panel) self._frame_set_helper() self.on_add_new_page(event=None) #Set the manager for the main panel self.fit_panel.set_manager(self) # List of windows used for the perspective self.perspective = [] self.perspective.append(self.fit_panel.window_name) self.result_frame = MDIFrame(self.parent, None, ResultPanel.window_caption, (220, 200)) self.result_panel = ResultPanel(parent=self.result_frame, manager=self) self.perspective.append(self.result_panel.window_name) #index number to create random model name self.index_model = 0 self.index_theory = 0 self.parent.Bind(EVT_SLICER_PANEL, self._on_slicer_event) self.parent.Bind(EVT_SLICER_PARS_UPDATE, self._onEVT_SLICER_PANEL) # CRUFT: EVT_FITTER_CHANGED is not None for bumps 0.7.5.9 and above if EVT_FITTER_CHANGED is not None: self.parent.Bind(EVT_FITTER_CHANGED, self.on_fitter_changed) self._set_fitter_label(bumps.options.FIT_CONFIG) #self.parent._mgr.Bind(wx.aui.EVT_AUI_PANE_CLOSE,self._onclearslicer) #Create reader when fitting panel are created self.state_reader = Reader(self.set_state) #append that reader to list of available reader loader = Loader() loader.associate_file_reader(".fitv", self.state_reader) #Send the fitting panel to guiframe self.mypanels.append(self.fit_panel) self.mypanels.append(self.result_panel) return self.mypanels
[docs] def clear_panel(self): """ """ self.fit_panel.clear_panel()
[docs] def delete_data(self, data): """ delete the given data from panel """ self.fit_panel.delete_data(data)
[docs] def set_data(self, data_list=None): """ receive a list of data to fit """ if data_list is None: data_list = [] selected_data_list = [] if self.batch_on: self.add_fit_page(data=data_list) else: if len(data_list) > MAX_NBR_DATA: dlg = DataDialog(data_list=data_list, nb_data=MAX_NBR_DATA) if dlg.ShowModal() == wx.ID_OK: selected_data_list = dlg.get_data() dlg.Destroy() else: selected_data_list = data_list try: group_id = wx.NewId() for data in selected_data_list: if data is not None: # 2D has no same group_id if data.__class__.__name__ == 'Data2D': group_id = wx.NewId() data.group_id = group_id if group_id not in data.list_group_id: data.list_group_id.append(group_id) self.add_fit_page(data=[data]) except: msg = "Fitting set_data: " + str(sys.exc_value) wx.PostEvent(self.parent, StatusEvent(status=msg, info="error"))
[docs] def set_theory(self, theory_list=None): """ """ #set the model state for a given theory_state: for item in theory_list: try: _, theory_state = item self.fit_panel.set_model_state(theory_state) except Exception: msg = "Fitting: cannot deal with the theory received" evt = StatusEvent(status=msg, info="error") logger.error("set_theory " + msg + "\n" + str(sys.exc_value)) wx.PostEvent(self.parent, evt)
[docs] def set_state(self, state=None, datainfo=None, format=None): """ Call-back method for the fit page state reader. This method is called when a .fitv/.svs file is loaded. : param state: PageState object : param datainfo: data """ if isinstance(state, PageState): state = state.clone() self.temp_state.append(state) elif isinstance(state, SimFitPageState): if self.fit_panel.sim_page is None: self.fit_panel.add_sim_page() self.fit_panel.sim_page.load_from_save_state(state) else: self.temp_state = [] # index to start with for a new set_state self.state_index = 0 # state file format self.sfile_ext = format self.on_set_state_helper(event=None)
[docs] def on_set_state_helper(self, event=None): """ Set_state_helper. This actually sets state after plotting data from state file. : event: FitStateUpdateEvent called by dataloader.plot_data from guiframe """ if len(self.temp_state) == 0: if self.state_index == 0 and len(self.mypanels) <= 0 \ and self.sfile_ext == '.svs': self.temp_state = [] self.state_index = 0 return try: # Load fitting state state = self.temp_state[self.state_index] #panel state should have model selection to set_state if state.formfactorcombobox is not None: #set state data = self.parent.create_gui_data(state.data) data.group_id = state.data.group_id self.parent.add_data(data_list={data.id: data}) wx.PostEvent(self.parent, NewPlotEvent(plot=data, title=data.title)) #need to be fix later make sure we are sendind guiframe.data #to panel state.data = data page = self.fit_panel.set_state(state) else: #just set data because set_state won't work data = self.parent.create_gui_data(state.data) data.group_id = state.data.group_id self.parent.add_data(data_list={data.id: data}) wx.PostEvent(self.parent, NewPlotEvent(plot=data, title=data.title)) page = self.add_fit_page([data]) caption = page.window_caption self.store_data(uid=page.uid, data_list=page.get_data_list(), caption=caption) self.mypanels.append(page) # get ready for the next set_state self.state_index += 1 #reset state variables to default when all set_state is finished. if len(self.temp_state) == self.state_index: self.temp_state = [] #self.state_index = 0 # Make sure the user sees the fitting panel after loading #self.parent.set_perspective(self.perspective) self.on_perspective(event=None) except: self.state_index = 0 self.temp_state = [] raise
[docs] def set_param2fit(self, uid, param2fit): """ Set the list of param names to fit for fitprobelm """ self.page_finder[uid].set_param2fit(param2fit)
[docs] def set_graph_id(self, uid, graph_id): """ Set graph_id for fitprobelm """ self.page_finder[uid].set_graph_id(graph_id)
[docs] def get_graph_id(self, uid): """ Set graph_id for fitprobelm """ return self.page_finder[uid].get_graph_id()
[docs] def save_fit_state(self, filepath, fitstate): """ save fit page state into file """ self.state_reader.write(filename=filepath, fitstate=fitstate)
[docs] def set_fit_weight(self, uid, flag, is2d=False, fid=None): """ Set the fit weights of a given page for all its data by default. If fid is provide then set the range only for the data with fid as id :param uid: id corresponding to a fit page :param fid: id corresponding to a fit problem (data, model) :param weight: current dy data """ # Note: this is used to set the data weights for the fit based on # the weight selection in the GUI. if uid in self.page_finder.keys(): self.page_finder[uid].set_weight(flag=flag, is2d=is2d)
[docs] def set_fit_range(self, uid, qmin, qmax, fid=None): """ Set the fitting range of a given page for all its data by default. If fid is provide then set the range only for the data with fid as id :param uid: id corresponding to a fit page :param fid: id corresponding to a fit problem (data, model) :param qmin: minimum value of the fit range :param qmax: maximum value of the fit range """ if uid in self.page_finder.keys(): self.page_finder[uid].set_range(qmin=qmin, qmax=qmax, fid=fid)
[docs] def schedule_for_fit(self, value=0, uid=None): """ Set the fit problem field to 0 or 1 to schedule that problem to fit. Schedule the specified fitproblem or get the fit problem related to the current page and set value. :param value: integer 0 or 1 :param uid: the id related to a page containing fitting information """ if uid in self.page_finder.keys(): self.page_finder[uid].schedule_tofit(value)
[docs] def get_page_finder(self): """ return self.page_finder used also by simfitpage.py """ return self.page_finder
[docs] def set_page_finder(self, modelname, names, values): """ Used by simfitpage.py to reset a parameter given the string constrainst. :param modelname: the name of the model for with the parameter has to reset :param value: can be a string in this case. :param names: the parameter name """ sim_page_id = self.sim_page.uid for uid, value in self.page_finder.iteritems(): if uid != sim_page_id and uid != self.batch_page.uid: model_list = value.get_model() model = model_list[0] if model.name == modelname: value.set_model_param(names, values) break
[docs] def split_string(self, item): """ receive a word containing dot and split it. used to split parameterset name into model name and parameter name example: :: parameterset (item) = M1.A Will return model_name = M1 , parameter name = A """ if item.find(".") >= 0: param_names = re.split(r"\.", item) model_name = param_names[0] ##Assume max len is 3; eg., M0.radius.width if len(param_names) == 3: param_name = param_names[1] + "." + param_names[2] else: param_name = param_names[1] return model_name, param_name
[docs] def on_bumps_options(self, event=None): """ Open the bumps options panel. """ show_fit_config(self.parent, help=self.on_help)
[docs] def on_fitter_changed(self, event): self._set_fitter_label(event.config)
def _set_fitter_label(self, config): self.fit_panel.parent.SetTitle(self.fit_panel.window_name + " - Active Fitting Optimizer: " + config.selected_name)
[docs] def on_help(self, algorithm_id): _TreeLocation = "user/sasgui/perspectives/fitting/optimizer.html" _anchor = "#fit-"+algorithm_id DocumentationWindow(self.parent, wx.ID_ANY, _TreeLocation, _anchor, "Optimizer Help")
[docs] def on_fit_results(self, event=None): """ Make the Fit Results panel visible. """ self.result_frame.Show() self.result_frame.Raise()
[docs] def on_gpu_options(self, event=None): """ Make the Fit Results panel visible. """ dialog = GpuOptions(None, wx.ID_ANY, "") dialog.Show()
[docs] def stop_fit(self, uid): """ Stop the fit """ if uid in self.fit_thread_list.keys(): calc_fit = self.fit_thread_list[uid] if calc_fit is not None and calc_fit.isrunning(): calc_fit.stop() msg = "Fit stop!" wx.PostEvent(self.parent, StatusEvent(status=msg, type="stop")) del self.fit_thread_list[uid] #set the fit button label of page when fit stop is trigger from #simultaneous fit pane sim_flag = self.sim_page is not None and uid == self.sim_page.uid batch_flag = self.batch_page is not None and uid == self.batch_page.uid if sim_flag or batch_flag: for uid, value in self.page_finder.iteritems(): if value.get_scheduled() == 1: if uid in self.fit_panel.opened_pages.keys(): panel = self.fit_panel.opened_pages[uid] panel._on_fit_complete()
[docs] def set_smearer(self, uid, smearer, fid, qmin=None, qmax=None, draw=True, enable_smearer=False): """ Get a smear object and store it to a fit problem of fid as id. If proper flag is enable , will plot the theory with smearing information. :param smearer: smear object to allow smearing data of id fid :param enable_smearer: Define whether or not all (data, model) contained in the structure of id uid will be smeared before fitting. :param qmin: the maximum value of the theory plotting range :param qmax: the maximum value of the theory plotting range :param draw: Determine if the theory needs to be plot """ if uid not in self.page_finder.keys(): return self.page_finder[uid].enable_smearing(flag=enable_smearer) self.page_finder[uid].set_smearer(smearer, fid=fid) if draw: ## draw model 1D with smeared data data = self.page_finder[uid].get_fit_data(fid=fid) if data is None: msg = "set_mearer requires at least data.\n" msg += "Got data = %s .\n" % str(data) return #raise ValueError, msg model = self.page_finder[uid].get_model(fid=fid) if model is None: return enable1D = issubclass(data.__class__, Data1D) enable2D = issubclass(data.__class__, Data2D) ## if user has already selected a model to plot ## redraw the model with data smeared smear = self.page_finder[uid].get_smearer(fid=fid) # compute weight for the current data weight = self.page_finder[uid].get_weight(fid=fid) self.draw_model(model=model, data=data, page_id=uid, smearer=smear, enable1D=enable1D, enable2D=enable2D, qmin=qmin, qmax=qmax, weight=weight)
[docs] def draw_model(self, model, page_id, data=None, smearer=None, enable1D=True, enable2D=False, state=None, fid=None, toggle_mode_on=False, qmin=None, qmax=None, update_chisqr=True, weight=None, source='model'): """ Draw model. :param model: the model to draw :param name: the name of the model to draw :param data: the data on which the model is based to be drawn :param description: model's description :param enable1D: if true enable drawing model 1D :param enable2D: if true enable drawing model 2D :param qmin: Range's minimum value to draw model :param qmax: Range's maximum value to draw model :param qstep: number of step to divide the x and y-axis :param update_chisqr: update chisqr [bool] """ #self.weight = weight if issubclass(data.__class__, Data1D) or not enable2D: ## draw model 1D with no loaded data self._draw_model1D(model=model, data=data, page_id=page_id, enable1D=enable1D, smearer=smearer, qmin=qmin, qmax=qmax, fid=fid, weight=weight, toggle_mode_on=toggle_mode_on, state=state, update_chisqr=update_chisqr, source=source) else: ## draw model 2D with no initial data self._draw_model2D(model=model, page_id=page_id, data=data, enable2D=enable2D, smearer=smearer, qmin=qmin, qmax=qmax, fid=fid, weight=weight, state=state, toggle_mode_on=toggle_mode_on, update_chisqr=update_chisqr, source=source)
[docs] def onFit(self, uid): """ Get series of data, model, associates parameters and range and send then to series of fitters. Fit data and model, display result to corresponding panels. :param uid: id related to the panel currently calling this fit function. """ if uid is None: raise RuntimeError("no page to fit") # Should never happen sim_page_uid = getattr(self.sim_page, 'uid', None) batch_page_uid = getattr(self.batch_page, 'uid', None) if uid == sim_page_uid: fit_type = 'simultaneous' elif uid == batch_page_uid: fit_type = 'combined_batch' else: fit_type = 'single' fitter_list = [] sim_fitter = None if fit_type == 'simultaneous': # for simultaneous fitting only one fitter is needed sim_fitter = Fit() sim_fitter.fitter_id = self.sim_page.uid fitter_list.append(sim_fitter) self.current_pg = None list_page_id = [] fit_id = 0 for page_id, page_info in self.page_finder.iteritems(): # For simulfit (uid give with None), do for-loop # if uid is specified (singlefit), do it only on the page. if page_id in (sim_page_uid, batch_page_uid): continue if fit_type == "single" and page_id != uid: continue try: if page_info.get_scheduled() == 1: page_info.nbr_residuals_computed = 0 page = self.fit_panel.get_page_by_id(page_id) self.set_fit_weight(uid=page.uid, flag=page.get_weight_flag(), is2d=page._is_2D()) if not page.param_toFit: msg = "No fitting parameters for %s" % page.window_caption evt = StatusEvent(status=msg, info="error", type="stop") wx.PostEvent(page.parent.parent, evt) return False if not page._update_paramv_on_fit(): msg = "Fitting range or parameter values are" msg += " invalid in %s" % \ page.window_caption evt = StatusEvent(status=msg, info="error", type="stop") wx.PostEvent(page.parent.parent, evt) return False pars = [str(element[1]) for element in page.param_toFit] fitproblem_list = page_info.values() for fitproblem in fitproblem_list: if sim_fitter is None: fitter = Fit() fitter.fitter_id = page_id fitter_list.append(fitter) else: fitter = sim_fitter self._add_problem_to_fit(fitproblem=fitproblem, pars=pars, fitter=fitter, fit_id=fit_id) fit_id += 1 list_page_id.append(page_id) page_info.clear_model_param() except KeyboardInterrupt: msg = "Fitting terminated" evt = StatusEvent(status=msg, info="info", type="stop") wx.PostEvent(self.parent, evt) return True except: raise msg = "Fitting error: %s" % str(sys.exc_value) evt = StatusEvent(status=msg, info="error", type="stop") wx.PostEvent(self.parent, evt) return False ## If a thread is already started, stop it #if self.calc_fitis not None and self.calc_fit.isrunning(): # self.calc_fit.stop() msg = "Fitting is in progress..." wx.PostEvent(self.parent, StatusEvent(status=msg, type="progress")) #Handler used to display fit message handler = ConsoleUpdate(parent=self.parent, manager=self, improvement_delta=0.1) # batch fit batch_inputs = {} batch_outputs = {} if fit_type == "simultaneous": page = self.sim_page elif fit_type == "combined_batch": page = self.batch_page else: page = self.fit_panel.get_page_by_id(uid) if page.batch_on: calc_fit = FitThread(handler=handler, fn=fitter_list, pars=pars, batch_inputs=batch_inputs, batch_outputs=batch_outputs, page_id=list_page_id, completefn=self._batch_fit_complete, reset_flag=self.batch_reset_flag) else: ## Perform more than 1 fit at the time calc_fit = FitThread(handler=handler, fn=fitter_list, batch_inputs=batch_inputs, batch_outputs=batch_outputs, page_id=list_page_id, updatefn=handler.update_fit, completefn=self._fit_completed) #self.fit_thread_list[current_page_id] = calc_fit self.fit_thread_list[uid] = calc_fit calc_fit.queue() calc_fit.ready(2.5) msg = "Fitting is in progress..." wx.PostEvent(self.parent, StatusEvent(status=msg, type="progress")) return True
[docs] def remove_plot(self, uid, fid=None, theory=False): """ remove model plot when a fit page is closed :param uid: the id related to the fitpage to close :param fid: the id of the fitproblem(data, model, range,etc) """ if uid not in self.page_finder.keys(): return fitproblemList = self.page_finder[uid].get_fit_problem(fid) for fitproblem in fitproblemList: data = fitproblem.get_fit_data() model = fitproblem.get_model() plot_id = None if model is not None: plot_id = data.id + model.name if theory: plot_id = data.id + model.name group_id = data.group_id wx.PostEvent(self.parent, NewPlotEvent(id=plot_id, group_id=group_id, action='remove'))
[docs] def store_data(self, uid, data_list=None, caption=None): """ Receive a list of data and store them ans well as a caption of the fit page where they come from. :param uid: if related to a fit page :param data_list: list of data to fit :param caption: caption of the window related to these data """ if data_list is None: data_list = [] self.page_finder[uid].set_fit_data(data=data_list) if caption is not None: self.page_finder[uid].set_fit_tab_caption(caption=caption)
[docs] def on_add_new_page(self, event=None): """ ask fit panel to create a new empty page """ try: page = self.fit_panel.add_empty_page() # add data associated to the page created if page is not None: evt = StatusEvent(status="Page Created", info="info") wx.PostEvent(self.parent, evt) else: msg = "Page was already Created" evt = StatusEvent(status=msg, info="warning") wx.PostEvent(self.parent, evt) except Exception: msg = "Creating Fit page: %s" % sys.exc_value wx.PostEvent(self.parent, StatusEvent(status=msg, info="error"))
[docs] def add_fit_page(self, data): """ given a data, ask to the fitting panel to create a new fitting page, get this page and store it into the page_finder of this plug-in :param data: is a list of data """ page = self.fit_panel.set_data(data) # page could be None when loading state files if page is None: return page #append Data1D to the panel containing its theory #if theory already plotted if page.uid in self.page_finder: data = page.get_data() theory_data = self.page_finder[page.uid].get_theory_data(data.id) if issubclass(data.__class__, Data2D): data.group_id = wx.NewId() if theory_data is not None: group_id = str(page.uid) + " Model1D" wx.PostEvent(self.parent, NewPlotEvent(group_id=group_id, action="delete")) self.parent.update_data(prev_data=theory_data, new_data=data) else: if theory_data is not None: group_id = str(page.uid) + " Model2D" data.group_id = theory_data.group_id wx.PostEvent(self.parent, NewPlotEvent(group_id=group_id, action="delete")) self.parent.update_data(prev_data=theory_data, new_data=data) self.store_data(uid=page.uid, data_list=page.get_data_list(), caption=page.window_caption) if self.sim_page is not None and not self.batch_on: self.sim_page.draw_page() if self.batch_page is not None and self.batch_on: self.batch_page.draw_page() return page
def _onEVT_SLICER_PANEL(self, event): """ receive and event telling to update a panel with a name starting with event.panel_name. this method update slicer panel for a given interactor. :param event: contains type of slicer , parameters for updating the panel and panel_name to find the slicer 's panel concerned. """ event.panel_name for item in self.parent.panels: name = event.panel_name if self.parent.panels[item].window_caption.startswith(name): self.parent.panels[item].set_slicer(event.type, event.params) #self.parent._mgr.Update() def _closed_fitpage(self, event): """ request fitpanel to close a given page when its unique data is removed from the plot. close fitpage only when the a loaded data is removed """ if event is None or event.data is None: return if hasattr(event.data, "is_data"): if not event.data.is_data or \ event.data.__class__.__name__ == "Data1D": self.fit_panel.close_page_with_data(event.data) def _reset_schedule_problem(self, value=0, uid=None): """ unschedule or schedule all fitproblem to be fit """ # case that uid is not specified if uid is None: for page_id in self.page_finder.keys(): self.page_finder[page_id].schedule_tofit(value) # when uid is given else: if uid in self.page_finder.keys(): self.page_finder[uid].schedule_tofit(value) def _add_problem_to_fit(self, fitproblem, pars, fitter, fit_id): """ Create and set fitter with series of data and model """ data = fitproblem.get_fit_data() model = fitproblem.get_model() smearer = fitproblem.get_smearer() qmin, qmax = fitproblem.get_range() #Extra list of parameters and their constraints listOfConstraint = [] param = fitproblem.get_model_param() if len(param) > 0: for item in param: ## check if constraint if item[0] is not None and item[1] is not None: listOfConstraint.append((item[0], item[1])) new_model = model fitter.set_model(new_model, fit_id, pars, data=data, constraints=listOfConstraint) fitter.set_data(data=data, id=fit_id, smearer=smearer, qmin=qmin, qmax=qmax) fitter.select_problem_for_fit(id=fit_id, value=1) def _onSelect(self, event): """ when Select data to fit a new page is created .Its reference is added to self.page_finder """ panel = self.plot_panel if panel is None: raise ValueError, "Fitting:_onSelect: NonType panel" Plugin.on_perspective(self, event=event) self.select_data(panel)
[docs] def select_data(self, panel): """ """ for plottable in panel.graph.plottables: if plottable.__class__.__name__ in ["Data1D", "Theory1D"]: data_id = panel.graph.selected_plottable if plottable == panel.plots[data_id]: data = plottable self.add_fit_page(data=[data]) return else: data = plottable self.add_fit_page(data=[data])
[docs] def update_fit(self, result=None, msg=""): """ """ print("update_fit result", result)
def _batch_fit_complete(self, result, pars, page_id, batch_outputs, batch_inputs, elapsed=None): """ Display fit result in batch :param result: list of objects received from fitters :param pars: list of fitted parameters names :param page_id: list of page ids which called fit function :param elapsed: time spent at the fitting level """ uid = page_id[0] if uid in self.fit_thread_list.keys(): del self.fit_thread_list[uid] wx.CallAfter(self._update_fit_button, page_id) t1 = time.time() str_time = time.strftime("%a, %d %b %Y %H:%M:%S ", time.localtime(t1)) msg = "Fit completed on %s \n" % str_time msg += "Duration time: %s s.\n" % str(elapsed) evt = StatusEvent(status=msg, info="info", type="stop") wx.PostEvent(self.parent, evt) if batch_outputs is None: batch_outputs = {} # format batch_outputs batch_outputs["Chi2"] = [] #Don't like these loops # Need to create dictionary of all fitted parameters # since the number of parameters can differ between each fit result for list_res in result: for res in list_res: model, data = res.inputs[0] if model is not None and hasattr(model, "model"): model = model.model #get all fittable parameters of the current model for param in model.getParamList(): if param not in batch_outputs.keys(): batch_outputs[param] = [] for param in model.getDispParamList(): if not model.is_fittable(param) and \ param in batch_outputs.keys(): del batch_outputs[param] # Add fitted parameters and their error for param in res.param_list: if param not in batch_outputs.keys(): batch_outputs[param] = [] err_param = "error on %s" % str(param) if err_param not in batch_inputs.keys(): batch_inputs[err_param] = [] msg = "" for list_res in result: for res in list_res: pid = res.fitter_id model, data = res.inputs[0] correct_result = False if model is not None and hasattr(model, "model"): model = model.model if data is not None and hasattr(data, "sas_data"): data = data.sas_data is_data2d = issubclass(data.__class__, Data2D) # Check consistency of arrays if not is_data2d: if len(res.theory) == len(res.index[res.index]) and \ len(res.index) == len(data.y): correct_result = True else: copy_data = deepcopy(data) new_theory = copy_data.data new_theory[res.index] = res.theory new_theory[res.index == False] = np.nan correct_result = True # Get all fittable parameters of the current model param_list = model.getParamList() for param in model.getDispParamList(): if '.' in param and param in param_list: # Ensure polydispersity results are displayed p1, p2 = param.split('.') if not model.is_fittable(p1) and not (p2 == 'width' and param in res.param_list)\ and param in param_list: param_list.remove(param) elif not model.is_fittable(param) and \ param in param_list: param_list.remove(param) if not correct_result or res.fitness is None or \ not np.isfinite(res.fitness) or \ np.any(res.pvec is None) or not \ np.all(np.isfinite(res.pvec)): data_name = str(None) if data is not None: data_name = str(data.name) model_name = str(None) if model is not None: model_name = str(model.name) msg += "Data %s and Model %s did not fit.\n" % (data_name, model_name) ERROR = np.NAN cell = BatchCell() cell.label = res.fitness cell.value = res.fitness batch_outputs["Chi2"].append(ERROR) for param in param_list: # Save value of fixed parameters if param not in res.param_list: batch_outputs[str(param)].append(ERROR) else: # Save only fitted values batch_outputs[param].append(ERROR) batch_inputs["error on %s" % str(param)].append(ERROR) else: # TODO: Why sometimes res.pvec comes with np.float64? # probably from scipy lmfit if res.pvec.__class__ == np.float64: res.pvec = [res.pvec] cell = BatchCell() cell.label = res.fitness cell.value = res.fitness batch_outputs["Chi2"].append(cell) # add parameters to batch_results for param in param_list: # save value of fixed parameters if param not in res.param_list: batch_outputs[str(param)].append(model.getParam(param)) else: index = res.param_list.index(param) #save only fitted values batch_outputs[param].append(res.pvec[index]) if res.stderr is not None and \ len(res.stderr) == len(res.param_list): item = res.stderr[index] batch_inputs["error on %s" % param].append(item) else: batch_inputs["error on %s" % param].append('-') model.setParam(param, res.pvec[index]) #fill the batch result with emtpy value if not in the current #model EMPTY = "-" for key in batch_outputs.keys(): if key not in param_list and key not in ["Chi2", "Data"]: batch_outputs[key].append(EMPTY) self.page_finder[pid].set_batch_result(batch_inputs=batch_inputs, batch_outputs=batch_outputs) cpage = self.fit_panel.get_page_by_id(pid) cpage._on_fit_complete() self.page_finder[pid][data.id].set_result(res) fitproblem = self.page_finder[pid][data.id] qmin, qmax = fitproblem.get_range() plot_result = False if correct_result: if not is_data2d: self._complete1D(x=data.x[res.index], y=res.theory, page_id=pid, elapsed=None, index=res.index, model=model, weight=None, fid=data.id, toggle_mode_on=False, state=None, data=data, update_chisqr=False, source='fit', plot_result=plot_result) else: self._complete2D(image=new_theory, data=data, model=model, page_id=pid, elapsed=None, index=res.index, qmin=qmin, qmax=qmax, fid=data.id, weight=None, toggle_mode_on=False, state=None, update_chisqr=False, source='fit', plot_result=plot_result) self.on_set_batch_result(page_id=pid, fid=data.id, batch_outputs=batch_outputs, batch_inputs=batch_inputs) evt = StatusEvent(status=msg, error="info", type="stop") wx.PostEvent(self.parent, evt) # Remove parameters that are not shown cpage = self.fit_panel.get_page_by_id(uid) tbatch_outputs = {} shownkeystr = cpage.get_copy_params() for key in batch_outputs.keys(): if key in ["Chi2", "Data"] or shownkeystr.count(key) > 0: tbatch_outputs[key] = batch_outputs[key] wx.CallAfter(self.parent.on_set_batch_result, tbatch_outputs, batch_inputs, self.sub_menu)
[docs] def on_set_batch_result(self, page_id, fid, batch_outputs, batch_inputs): """ """ pid = page_id if fid not in self.page_finder[pid]: return fitproblem = self.page_finder[pid][fid] index = self.page_finder[pid].nbr_residuals_computed - 1 residuals = fitproblem.get_residuals() theory_data = fitproblem.get_theory_data() data = fitproblem.get_fit_data() model = fitproblem.get_model() #fill batch result information if "Data" not in batch_outputs.keys(): batch_outputs["Data"] = [] cell = BatchCell() cell.label = data.name cell.value = index if theory_data is not None: #Suucessful fit theory_data.id = wx.NewId() theory_data.name = model.name + "[%s]" % str(data.name) if issubclass(theory_data.__class__, Data2D): group_id = wx.NewId() theory_data.group_id = group_id if group_id not in theory_data.list_group_id: theory_data.list_group_id.append(group_id) try: # associate residuals plot if issubclass(residuals.__class__, Data2D): group_id = wx.NewId() residuals.group_id = group_id if group_id not in residuals.list_group_id: residuals.list_group_id.append(group_id) batch_outputs["Chi2"][index].object = [residuals] except: pass cell.object = [data, theory_data] batch_outputs["Data"].append(cell) for key, value in data.meta_data.iteritems(): if key not in batch_inputs.keys(): batch_inputs[key] = [] #if key.lower().strip() != "loader": batch_inputs[key].append(value) param = "temperature" if hasattr(data.sample, param): if param not in batch_inputs.keys(): batch_inputs[param] = [] batch_inputs[param].append(data.sample.temperature)
def _fit_completed(self, result, page_id, batch_outputs, batch_inputs=None, pars=None, elapsed=None): """ Display result of the fit on related panel(s). :param result: list of object generated when fit ends :param pars: list of names of parameters fitted :param page_id: list of page ids which called fit function :param elapsed: time spent at the fitting level """ t1 = time.time() str_time = time.strftime("%a, %d %b %Y %H:%M:%S ", time.localtime(t1)) msg = "Fit completed on %s \n" % str_time msg += "Duration time: %s s.\n" % str(elapsed) evt = StatusEvent(status=msg, info="info", type="stop") wx.PostEvent(self.parent, evt) wx.PostEvent(self.result_panel, PlotResultEvent(result=result)) wx.CallAfter(self._update_fit_button, page_id) result = result[0] self.fit_thread_list = {} if page_id is None: page_id = [] ## fit more than 1 model at the same time try: index = 0 # Update potential simfit page(s) if self.sim_page is not None: self.sim_page._on_fit_complete() if self.batch_page: self.batch_page._on_fit_complete() # Update all fit pages for uid in page_id: res = result[index] fit_msg = res.mesg if res.fitness is None or \ not np.isfinite(res.fitness) or \ np.any(res.pvec is None) or \ not np.all(np.isfinite(res.pvec)): fit_msg += "\nFitting did not converge!!!" wx.CallAfter(self._update_fit_button, page_id) else: #set the panel when fit result are float not list if res.pvec.__class__ == np.float64: pvec = [res.pvec] else: pvec = res.pvec if res.stderr.__class__ == np.float64: stderr = [res.stderr] else: stderr = res.stderr cpage = self.fit_panel.get_page_by_id(uid) # Make sure we got all results #(CallAfter is important to MAC) try: #if res is not None: wx.CallAfter(cpage.onsetValues, res.fitness, res.param_list, pvec, stderr) index += 1 wx.CallAfter(cpage._on_fit_complete) except KeyboardInterrupt: fit_msg += "\nSingular point: Fitting stopped." except Exception: fit_msg += "\nSingular point: Fitting error occurred." if fit_msg: evt = StatusEvent(status=fit_msg, info="warning", type="stop") wx.PostEvent(self.parent, evt) except Exception: msg = ("Fit completed but the following error occurred: %s" % sys.exc_value) #msg = "\n".join((traceback.format_exc(), msg)) evt = StatusEvent(status=msg, info="warning", type="stop") wx.PostEvent(self.parent, evt) def _update_fit_button(self, page_id): """ Update Fit button when fit stopped : parameter page_id: fitpage where the button is """ if page_id.__class__.__name__ != 'list': page_id = [page_id] for uid in page_id: page = self.fit_panel.get_page_by_id(uid) page._on_fit_complete() def _on_show_panel(self, event): """ """ pass
[docs] def on_reset_batch_flag(self, event): """ Set batch_reset_flag """ event.Skip() if self.menu1 is None: return menu_item = self.menu1.FindItemById(self.id_reset_flag) flag = menu_item.IsChecked() if not flag: menu_item.Check(False) self.batch_reset_flag = True else: menu_item.Check(True) self.batch_reset_flag = False ## post a message to status bar msg = "Set Chain Fitting: %s" % str(not self.batch_reset_flag) wx.PostEvent(self.parent, StatusEvent(status=msg))
def _on_slicer_event(self, event): """ Receive a panel as event and send it to guiframe :param event: event containing a panel """ if event.panel is not None: self.slicer_panels.append(event.panel) # Set group ID if available event_id = self.parent.popup_panel(event.panel) event.panel.uid = event_id self.mypanels.append(event.panel) def _onclearslicer(self, event): """ Clear the boxslicer when close the panel associate with this slicer """ name = event.GetEventObject().frame.GetTitle() for panel in self.slicer_panels: if panel.window_caption == name: for item in self.parent.panels: if hasattr(self.parent.panels[item], "uid"): if self.parent.panels[item].uid == panel.base.uid: self.parent.panels[item].onClearSlicer(event) #self.parent._mgr.Update() break break def _on_model_panel(self, evt): """ react to model selection on any combo box or model menu.plot the model :param evt: wx.combobox event """ model = evt.model uid = evt.uid qmin = evt.qmin qmax = evt.qmax caption = evt.caption enable_smearer = evt.enable_smearer if model is None: return if uid not in self.page_finder.keys(): return # save the name containing the data name with the appropriate model self.page_finder[uid].set_model(model) self.page_finder[uid].enable_smearing(enable_smearer) self.page_finder[uid].set_range(qmin=qmin, qmax=qmax) self.page_finder[uid].set_fit_tab_caption(caption=caption) if self.sim_page is not None and not self.batch_on: self.sim_page.draw_page() if self.batch_page is not None and self.batch_on: self.batch_page.draw_page() def _update1D(self, x, output): """ Update the output of plotting model 1D """ msg = "Plot updating ... " wx.PostEvent(self.parent, StatusEvent(status=msg, type="update"))
[docs] def create_theory_1D(self, x, y, page_id, model, data, state, data_description, data_id, dy=None): """ Create a theory object associate with an existing Data1D and add it to the data manager. @param x: x-values of the data @param y: y_values of the data @param page_id: fit page ID @param model: model used for fitting @param data: Data1D object to create the theory for @param state: model state @param data_description: title to use in the data manager @param data_id: unique data ID """ new_plot = Data1D(x=x, y=y) if dy is None: new_plot.is_data = False new_plot.dy = np.zeros(len(y)) # If this is a theory curve, pick the proper symbol to make it a curve new_plot.symbol = GUIFRAME_ID.CURVE_SYMBOL_NUM else: new_plot.is_data = True new_plot.dy = dy new_plot.interactive = True new_plot.dx = None new_plot.dxl = None new_plot.dxw = None _yaxis, _yunit = data.get_yaxis() _xaxis, _xunit = data.get_xaxis() new_plot.title = data.name new_plot.group_id = data.group_id if new_plot.group_id is None: new_plot.group_id = data.group_id new_plot.id = data_id # Find if this theory was already plotted and replace that plot given # the same id self.page_finder[page_id].get_theory_data(fid=data.id) if data.is_data: data_name = str(data.name) else: data_name = str(model.__class__.__name__) new_plot.name = data_description + " [" + data_name + "]" new_plot.xaxis(_xaxis, _xunit) new_plot.yaxis(_yaxis, _yunit) self.page_finder[page_id].set_theory_data(data=new_plot, fid=data.id) self.parent.update_theory(data_id=data.id, theory=new_plot, state=state) return new_plot
def _complete1D(self, x, y, page_id, elapsed, index, model, weight=None, fid=None, toggle_mode_on=False, state=None, data=None, update_chisqr=True, source='model', plot_result=True, unsmeared_model=None, unsmeared_data=None, unsmeared_error=None, sq_model=None, pq_model=None): """ Complete plotting 1D data @param unsmeared_model: fit model, without smearing @param unsmeared_data: data, rescaled to unsmeared model @param unsmeared_error: data error, rescaled to unsmeared model """ number_finite = np.count_nonzero(np.isfinite(y)) np.nan_to_num(y) new_plot = self.create_theory_1D(x, y, page_id, model, data, state, data_description=model.name, data_id=str(page_id) + " " + data.name) plots_to_update = [] # List of plottables that have changed since last calculation # Create the new theories if unsmeared_model is not None: unsmeared_model_plot = self.create_theory_1D(x, unsmeared_model, page_id, model, data, state, data_description=model.name + " unsmeared", data_id=str(page_id) + " " + data.name + " unsmeared") plots_to_update.append(unsmeared_model_plot) if unsmeared_data is not None and unsmeared_error is not None: unsmeared_data_plot = self.create_theory_1D(x, unsmeared_data, page_id, model, data, state, data_description="Data unsmeared", data_id="Data " + data.name + " unsmeared", dy=unsmeared_error) plots_to_update.append(unsmeared_data_plot) if sq_model is not None and pq_model is not None: sq_id = str(page_id) + " " + data.name + " S(q)" sq_plot = self.create_theory_1D(x, sq_model, page_id, model, data, state, data_description=model.name + " S(q)", data_id=sq_id) plots_to_update.append(sq_plot) pq_id = str(page_id) + " " + data.name + " P(q)" pq_plot = self.create_theory_1D(x, pq_model, page_id, model, data, state, data_description=model.name + " P(q)", data_id=pq_id) plots_to_update.append(pq_plot) # Update the P(Q), S(Q) and unsmeared theory plots if they exist wx.PostEvent(self.parent, NewPlotEvent(plots=plots_to_update, action='update')) current_pg = self.fit_panel.get_page_by_id(page_id) title = new_plot.title batch_on = self.fit_panel.get_page_by_id(page_id).batch_on if not batch_on: wx.PostEvent(self.parent, NewPlotEvent(plot=new_plot, title=str(title))) elif plot_result: top_data_id = self.fit_panel.get_page_by_id(page_id).data.id if data.id == top_data_id: wx.PostEvent(self.parent, NewPlotEvent(plot=new_plot, title=str(title))) caption = current_pg.window_caption self.page_finder[page_id].set_fit_tab_caption(caption=caption) self.page_finder[page_id].set_theory_data(data=new_plot, fid=data.id) if toggle_mode_on: wx.PostEvent(self.parent, NewPlotEvent(group_id=str(page_id) + " Model2D", action="Hide")) else: if update_chisqr: output = self._cal_chisqr(data=data, fid=fid, weight=weight, page_id=page_id, index=index) wx.PostEvent(current_pg, Chi2UpdateEvent(output=output)) else: self._plot_residuals(page_id=page_id, data=data, fid=fid, index=index, weight=weight) if not number_finite: logger.error("Using the present parameters the model does not return any finite value. ") msg = "Computing Error: Model did not return any finite value." wx.PostEvent(self.parent, StatusEvent(status=msg, info="error")) else: msg = "Computation completed!" if number_finite != y.size: msg += ' PROBLEM: For some Q values the model returns non finite intensities!' logger.error("For some Q values the model returns non finite intensities.") wx.PostEvent(self.parent, StatusEvent(status=msg, type="stop")) def _calc_exception(self, etype, value, tb): """ Handle exception from calculator by posting it as an error. """ logger.error("".join(traceback.format_exception(etype, value, tb))) msg = traceback.format_exception(etype, value, tb, limit=1) evt = StatusEvent(status="".join(msg), type="stop", info="error") wx.PostEvent(self.parent, evt) def _update2D(self, output, time=None): """ Update the output of plotting model """ msg = "Plot updating ... " wx.PostEvent(self.parent, StatusEvent(msg, type="update")) def _complete2D(self, image, data, model, page_id, elapsed, index, qmin, qmax, fid=None, weight=None, toggle_mode_on=False, state=None, update_chisqr=True, source='model', plot_result=True): """ Complete get the result of modelthread and create model 2D that can be plot. """ number_finite = np.count_nonzero(np.isfinite(image)) np.nan_to_num(image) new_plot = Data2D(image=image, err_image=data.err_data) new_plot.name = model.name + '2d' new_plot.title = "Analytical model 2D " new_plot.id = str(page_id) + " " + data.name new_plot.group_id = str(page_id) + " Model2D" new_plot.detector = data.detector new_plot.source = data.source new_plot.is_data = False new_plot.qx_data = data.qx_data new_plot.qy_data = data.qy_data new_plot.q_data = data.q_data new_plot.mask = data.mask ## plot boundaries new_plot.ymin = data.ymin new_plot.ymax = data.ymax new_plot.xmin = data.xmin new_plot.xmax = data.xmax title = data.title new_plot.is_data = False if data.is_data: data_name = str(data.name) else: data_name = str(model.__class__.__name__) + '2d' if len(title) > 1: new_plot.title = "Model2D for %s " % model.name + data_name new_plot.name = model.name + " [" + \ data_name + "]" theory_data = deepcopy(new_plot) self.page_finder[page_id].set_theory_data(data=theory_data, fid=data.id) self.parent.update_theory(data_id=data.id, theory=new_plot, state=state) current_pg = self.fit_panel.get_page_by_id(page_id) title = new_plot.title if not source == 'fit' and plot_result: wx.PostEvent(self.parent, NewPlotEvent(plot=new_plot, title=title)) if toggle_mode_on: wx.PostEvent(self.parent, NewPlotEvent(group_id=str(page_id) + " Model1D", action="Hide")) else: # Chisqr in fitpage if update_chisqr: output = self._cal_chisqr(data=data, weight=weight, fid=fid, page_id=page_id, index=index) wx.PostEvent(current_pg, Chi2UpdateEvent(output=output)) else: self._plot_residuals(page_id=page_id, data=data, fid=fid, index=index, weight=weight) if not number_finite: logger.error("Using the present parameters the model does not return any finite value. ") msg = "Computing Error: Model did not return any finite value." wx.PostEvent(self.parent, StatusEvent(status=msg, info="error")) else: msg = "Computation completed!" if number_finite != image.size: msg += ' PROBLEM: For some Qx,Qy values the model returns non finite intensities!' logger.error("For some Qx,Qy values the model returns non finite intensities.") wx.PostEvent(self.parent, StatusEvent(status=msg, type="stop")) def _draw_model2D(self, model, page_id, qmin, qmax, data=None, smearer=None, description=None, enable2D=False, state=None, fid=None, weight=None, toggle_mode_on=False, update_chisqr=True, source='model'): """ draw model in 2D :param model: instance of the model to draw :param description: the description of the model :param enable2D: when True allows to draw model 2D :param qmin: the minimum value to draw model 2D :param qmax: the maximum value to draw model 2D :param qstep: the number of division of Qx and Qy of the model to draw """ if not enable2D: return None try: ## If a thread is already started, stop it if (self.calc_2D is not None) and self.calc_2D.isrunning(): self.calc_2D.stop() ## stop just raises a flag to tell the thread to kill ## itself -- see the fix in Calc1D implemented to fix ## an actual problem. Seems the fix should also go here ## and may be the cause of other noted instabilities ## ## -PDB August 12, 2014 while self.calc_2D.isrunning(): time.sleep(0.1) self.calc_2D = Calc2D(model=model, data=data, page_id=page_id, smearer=smearer, qmin=qmin, qmax=qmax, weight=weight, fid=fid, toggle_mode_on=toggle_mode_on, state=state, completefn=self._complete2D, update_chisqr=update_chisqr, exception_handler=self._calc_exception, source=source) self.calc_2D.queue() except: raise def _draw_model1D(self, model, page_id, data, qmin, qmax, smearer=None, state=None, weight=None, fid=None, toggle_mode_on=False, update_chisqr=True, source='model', enable1D=True): """ Draw model 1D from loaded data1D :param data: loaded data :param model: the model to plot """ if not enable1D: return try: ## If a thread is already started, stop it if (self.calc_1D is not None) and self.calc_1D.isrunning(): self.calc_1D.stop() ## stop just raises the flag -- the thread is supposed to ## then kill itself but cannot. Paul Kienzle came up with ## this fix to prevent threads from stepping on each other ## which was causing a simple custom plugin model to crash ##Sasview. ## We still don't know why the fit sometimes lauched a second ## thread -- something which should also be investigated. ## The thread approach was implemented in order to be able ## to lauch a computation in a separate thread from the GUI so ## that the GUI can still respond to user input including ## a request to stop the computation. ## It seems thus that the whole thread approach used here ## May need rethinking ## ## -PDB August 12, 2014 while self.calc_1D.isrunning(): time.sleep(0.1) self.calc_1D = Calc1D(data=data, model=model, page_id=page_id, qmin=qmin, qmax=qmax, smearer=smearer, state=state, weight=weight, fid=fid, toggle_mode_on=toggle_mode_on, completefn=self._complete1D, #updatefn = self._update1D, update_chisqr=update_chisqr, exception_handler=self._calc_exception, source=source) self.calc_1D.queue() except: msg = " Error occurred when drawing %s Model 1D: " % model.name msg += " %s" % sys.exc_value wx.PostEvent(self.parent, StatusEvent(status=msg)) def _cal_chisqr(self, page_id, data, weight, fid=None, index=None): """ Get handy Chisqr using the output from draw1D and 2D, instead of calling expansive CalcChisqr in guithread """ try: data_copy = deepcopy(data) except: return # default chisqr chisqr = None #to compute chisq make sure data has valid data # return None if data is None if not check_data_validity(data_copy) or data_copy is None: return chisqr # Get data: data I, theory I, and data dI in order if data_copy.__class__.__name__ == "Data2D": if index is None: index = np.ones(len(data_copy.data), dtype=bool) if weight is not None: data_copy.err_data = weight # get rid of zero error points index = index & (data_copy.err_data != 0) index = index & (np.isfinite(data_copy.data)) fn = data_copy.data[index] theory_data = self.page_finder[page_id].get_theory_data(fid=data_copy.id) if theory_data is None: return chisqr gn = theory_data.data[index] en = data_copy.err_data[index] else: # 1 d theory from model_thread is only in the range of index if index is None: index = np.ones(len(data_copy.y), dtype=bool) if weight is not None: data_copy.dy = weight if data_copy.dy is None or data_copy.dy == []: dy = np.ones(len(data_copy.y)) else: ## Set consistently w/AbstractFitengine: # But this should be corrected later. dy = deepcopy(data_copy.dy) dy[dy == 0] = 1 fn = data_copy.y[index] theory_data = self.page_finder[page_id].get_theory_data(fid=data_copy.id) if theory_data is None: return chisqr gn = theory_data.y en = dy[index] # residual try: res = (fn - gn) / en except ValueError: print("Unmatch lengths %s, %s, %s" % (len(fn), len(gn), len(en))) return residuals = res[np.isfinite(res)] # get chisqr only w/finite chisqr = np.average(residuals * residuals) self._plot_residuals(page_id=page_id, data=data_copy, fid=fid, weight=weight, index=index) return chisqr def _plot_residuals(self, page_id, weight, fid=None, data=None, index=None): """ Plot the residuals :param data: data :param index: index array (bool) : Note: this is different from the residuals in cal_chisqr() """ data_copy = deepcopy(data) # Get data: data I, theory I, and data dI in order if data_copy.__class__.__name__ == "Data2D": # build residuals residuals = Data2D() #residuals.copy_from_datainfo(data) # Not for trunk the line below, instead use the line above data_copy.clone_without_data(len(data_copy.data), residuals) residuals.data = None fn = data_copy.data theory_data = self.page_finder[page_id].get_theory_data(fid=data_copy.id) gn = theory_data.data if weight is None: en = data_copy.err_data else: en = weight residuals.data = (fn - gn) / en residuals.qx_data = data_copy.qx_data residuals.qy_data = data_copy.qy_data residuals.q_data = data_copy.q_data residuals.err_data = np.ones(len(residuals.data)) residuals.xmin = min(residuals.qx_data) residuals.xmax = max(residuals.qx_data) residuals.ymin = min(residuals.qy_data) residuals.ymax = max(residuals.qy_data) residuals.q_data = data_copy.q_data residuals.mask = data_copy.mask residuals.scale = 'linear' # check the lengths if len(residuals.data) != len(residuals.q_data): return else: # 1 d theory from model_thread is only in the range of index if data_copy.dy is None or data_copy.dy == []: dy = np.ones(len(data_copy.y)) else: if weight is None: dy = np.ones(len(data_copy.y)) ## Set consitently w/AbstractFitengine: ## But this should be corrected later. else: dy = weight dy[dy == 0] = 1 fn = data_copy.y[index] theory_data = self.page_finder[page_id].get_theory_data(fid=data_copy.id) gn = theory_data.y en = dy[index] # build residuals residuals = Data1D() try: residuals.y = (fn - gn) / en except: msg = "ResidualPlot Error: different # of data points in theory" wx.PostEvent(self.parent, StatusEvent(status=msg, info="error")) residuals.y = (fn - gn[index]) / en residuals.x = data_copy.x[index] residuals.dy = np.ones(len(residuals.y)) residuals.dx = None residuals.dxl = None residuals.dxw = None residuals.ytransform = 'y' # For latter scale changes residuals.xaxis('\\rm{Q} ', 'A^{-1}') residuals.yaxis('\\rm{Residuals} ', 'normalized') theory_name = str(theory_data.name.split()[0]) new_plot = residuals new_plot.name = "Residuals for " + str(theory_name) + "[" + \ str(data.name) + "]" ## allow to highlight data when plotted new_plot.interactive = True ## when 2 data have the same id override the 1 st plotted new_plot.id = "res" + str(data_copy.id) + str(theory_name) ##group_id specify on which panel to plot this data group_id = self.page_finder[page_id].get_graph_id() if group_id is None: group_id = data.group_id new_plot.group_id = "res" + str(group_id) #new_plot.is_data = True ##post data to plot title = new_plot.name self.page_finder[page_id].set_residuals(residuals=new_plot, fid=data.id) self.parent.update_theory(data_id=data.id, theory=new_plot) batch_on = self.fit_panel.get_page_by_id(page_id).batch_on if not batch_on: wx.PostEvent(self.parent, NewPlotEvent(plot=new_plot, title=title))