Example #1
0
 def normalize_selection(self):
     '''
     Normalize the current selection.
     '''
     self.selection_handler.create_selection()
     if len(self.selection_handler.selected_indices)==1:
         column_index = self.selection_handler.selected_indices[0][1]
         if self.AS_PANDAS_DATAFRAME:
             column_name = self.data_frame.columns[column_index]
             x = self.data_frame[column_name]
             x = x - x.mean()
             x = x/x.std()
             self.data_frame[column_name]=x
             self.table_model = DataFrameModel(
                 self.data_frame,
                 editable=True,
                 horizontal_headers=self.data_frame.columns
             )
         else:
             x = self.table[:,column_index]
             x = x - x.mean()
             x = x/x.std()
             self.table[:,column_index] = x
             self.table_model = TableModel(self.table, editable=True,
                                           horizontal_headers=self.headers)
     self.selection_handler.flush()
Example #2
0
 def normalize_selection(self):
     '''
     Normalize the current selection.
     '''
     self.selection_handler.create_selection()
     if len(self.selection_handler.selected_indices) == 1:
         column_index = self.selection_handler.selected_indices[0][1]
         if self.AS_PANDAS_DATAFRAME:
             column_name = self.data_frame.columns[column_index]
             x = self.data_frame[column_name]
             x = x - x.mean()
             x = x / x.std()
             self.data_frame[column_name] = x
             self.table_model = DataFrameModel(
                 self.data_frame,
                 editable=True,
                 horizontal_headers=self.data_frame.columns)
         else:
             x = self.table[:, column_index]
             x = x - x.mean()
             x = x / x.std()
             self.table[:, column_index] = x
             self.table_model = TableModel(self.table,
                                           editable=True,
                                           horizontal_headers=self.headers)
     self.selection_handler.flush()
Example #3
0
 def redraw_tablemodel(self):
     '''
     Called when the tabular view needs to be redrawn when the changes to the
     table are programmatic, not through the UI.
     '''
     
     if self.AS_PANDAS_DATAFRAME:
         self.table_model = DataFrameModel(
             self.data_frame, editable= True,
             horizontal_headers=self.data_frame.columns
         )
     else:
         self.table_model = TableModel(self.table, editable=True,
                                       horizontal_headers=self.headers)
Example #4
0
    def redraw_tablemodel(self):
        '''
        Called when the tabular view needs to be redrawn when the changes to the
        table are programmatic, not through the UI.
        '''

        if self.AS_PANDAS_DATAFRAME:
            self.table_model = DataFrameModel(
                self.data_frame,
                editable=True,
                horizontal_headers=self.data_frame.columns)
        else:
            self.table_model = TableModel(self.table,
                                          editable=True,
                                          horizontal_headers=self.headers)
Example #5
0
    def map_to_unique(self):
        '''
        USed to map a set of items onto unique integers,
        and change tabular display accordingly.
        '''
        self.selection_handler.create_selection()

        if len(self.selection_handler.selected_indices) == 1:
            # currently works only for a single coulmn
            self.unique_map = {}
            column = self.selection_handler.selected_indices[0][1]
            column_name = self.data_frame.columns[column]
            data = DataFrame(self.data_frame[column_name])
            m = 0

            gb = data.groupby([column_name])

            for key in gb.indices.keys():
                self.unique_map[key] = m
                m += 1
            new_indices = data[column_name]

            for i in range(len(new_indices)):
                key = new_indices[i]
                new_indices[i] = self.unique_map[key]
            self.data_frame[column_name] = new_indices

            self.table_model = DataFrameModel(
                self.data_frame,
                editable=True,
                horizontal_headers=self.data_frame.columns,
            )

        self.selection_handler.flush()
Example #6
0
    def _table_model_default(self):
        '''
        The default TableModel instance corresponding to the array of zeros
        '''

        tblmodel = TableModel(self.table, editable=True)
        return tblmodel
Example #7
0
 def map_to_unique(self):
     '''
     USed to map a set of items onto unique integers,
     and change tabular display accordingly.
     '''
     self.selection_handler.create_selection()
     
     if len(self.selection_handler.selected_indices)==1:
         # currently works only for a single coulmn
         self.unique_map = {}
         column = self.selection_handler.selected_indices[0][1]
         column_name = self.data_frame.columns[column]
         data = DataFrame(self.data_frame[column_name])
         m = 0
         
         gb = data.groupby([column_name])
         
         for key in gb.indices.keys():
             self.unique_map[key] = m
             m += 1
         new_indices = data[column_name]
         
         for i in range(len(new_indices)):
             key = new_indices[i]
             new_indices[i] = self.unique_map[key]
         self.data_frame[column_name] = new_indices
         
         self.table_model = DataFrameModel(
             self.data_frame,
             editable=True,
             horizontal_headers=self.data_frame.columns,
         )
     
     self.selection_handler.flush()
Example #8
0
 def _filename_changed(self, new):
     '''
     Executes whenever a file is loaded into the view.
     '''
     if self.AS_PANDAS_DATAFRAME:
         self.data_frame = read_csv(self.filename)
         self.table_model = DataFrameModel(
             self.data_frame,
             editable=True,
             horizontal_headers=self.data_frame.columns,
         )
     else:
         self.table = np.genfromtxt(self.filename, delimiter=',', skip_header=1)
         csv_reader= csv.reader(file(self.filename))
         self.headers = csv_reader.next()
         del csv_reader
         self.table_model = TableModel(self.table, editable=True,
                                  horizontal_headers=self.headers)
Example #9
0
 def _filename_changed(self, new):
     '''
     Executes whenever a file is loaded into the view.
     '''
     if self.AS_PANDAS_DATAFRAME:
         self.data_frame = read_csv(self.filename)
         self.table_model = DataFrameModel(
             self.data_frame,
             editable=True,
             horizontal_headers=self.data_frame.columns,
         )
     else:
         self.table = np.genfromtxt(self.filename,
                                    delimiter=',',
                                    skip_header=1)
         csv_reader = csv.reader(file(self.filename))
         self.headers = csv_reader.next()
         del csv_reader
         self.table_model = TableModel(self.table,
                                       editable=True,
                                       horizontal_headers=self.headers)
Example #10
0
class CsvModel(HasTraits):
    '''
    The object that is passed to the MainWindow of the enaml view. This
    represents a general model for different plots in the csv editor.
    '''

    # The pandas data frame associated with the current instance of the editor.
    data_frame = Instance(DataFrame, ())

    # Lenght of the selected column
    column_length = Int

    # Number of unique items in a selection
    unique_items_nos = Int

    # the unique mapping from a list of items to a list of integers
    unique_map = Dict

    # The .csv file to be opened.
    filename = File

    # The filename of the file to be saved
    save_filename = String

    # The numpy array associated with the data in the file
    table = Array

    # The headers of the csv data
    headers = List

    # The mean of the 2-D block of the array, represented by the current
    # selection
    selection_mean = Float

    # The variance of the 2-D block of the array, represented by the current
    # selection
    selection_var = Float

    # The standard deviation of the 2-D block of the array, represented by the
    # current selection
    selection_std = Float

    # The sum of the current selection
    selection_sum = Float

    # The minimum value from the current selection
    selection_min = Float

    # The maximum value from the current selection
    selection_max = Float

    # The covariance between a pair of rows or columns
    # NOT IMPLEMENTED
    selection_cov = Float(0)

    # The TableModel instance to be passed to the item_model attribute of the
    # TableView
    table_model = Instance(TableModel, ())

    # class for handling scripts
    script_handler = Instance(ScriptHandler)

    # class for handling the workspace generated by the script
    workspace_handler = Any

    # Class for handling selections from the TableView
    selection_handler = Instance(SelectionHandler)

    # If the file should be imported as a Pandas dataframe
    AS_PANDAS_DATAFRAME = Bool(False)

    def _workspace_handler_default(self):
        return WorkspaceHandler({})

    def _script_handler_default(self):
        return ScriptHandler()

    def _selection_handler_default(self):
        return SelectionHandler()

    def _table_default(self):
        '''
        The default array in the TableView is a 100x100 array of zeros
        '''
        if not self.AS_PANDAS_DATAFRAME:
            x = np.zeros((100, 100))
            return x
        else:
            return None

    def _data_frame_default(self):
        if self.AS_PANDAS_DATAFRAME:
            return DataFrame()
        else:
            return None

    def _table_model_default(self):
        '''
        The default TableModel instance corresponding to the array of zeros
        '''

        tblmodel = TableModel(self.table, editable=True)
        return tblmodel

    def _image_plot_default(self):
        '''
        Default chaco plot object for the image plot.
        '''

        self.img_plotdata = ArrayPlotData(imagedata=self.table)
        p = Plot(self.img_plotdata)
        p.img_plot('imagedata')
        return p

    def _hist_plot_default(self):
        '''
        Default chaco plot object for the histogram.
        '''

        h = np.histogram(self.table, 10)[0]
        self.hist_plotdata = ArrayPlotData(x=h)
        p = Plot(self.hist_plotdata)
        p.plot('x', type='bar', color='auto', bar_width=0.3)
        return p

    def _filename_changed(self, new):
        '''
        Executes whenever a file is loaded into the view.
        '''
        if self.AS_PANDAS_DATAFRAME:
            self.data_frame = read_csv(self.filename)
            self.table_model = DataFrameModel(
                self.data_frame,
                editable=True,
                horizontal_headers=self.data_frame.columns,
            )
        else:
            self.table = np.genfromtxt(self.filename,
                                       delimiter=',',
                                       skip_header=1)
            csv_reader = csv.reader(file(self.filename))
            self.headers = csv_reader.next()
            del csv_reader
            self.table_model = TableModel(self.table,
                                          editable=True,
                                          horizontal_headers=self.headers)

    def redraw_tablemodel(self):
        '''
        Called when the tabular view needs to be redrawn when the changes to the
        table are programmatic, not through the UI.
        '''

        if self.AS_PANDAS_DATAFRAME:
            self.table_model = DataFrameModel(
                self.data_frame,
                editable=True,
                horizontal_headers=self.data_frame.columns)
        else:
            self.table_model = TableModel(self.table,
                                          editable=True,
                                          horizontal_headers=self.headers)

    def _pca_default(self):
        return PCA(2, whiten=True)

    def create_plot_properties():
        pass

    def use_selection_preview(self):
        pass

    def use_selection_histogram(self):
        pass

    def use_selection_imageplot(self):
        pass

    def use_selection_pcaplot(self):
        pass

    def calculate_selection_params(self):
        '''
        Called to calculate statistics of a selection.
        '''
        self.selection_handler.create_selection()
        tuple_list = self.selection_handler.selected_indices

        if self.AS_PANDAS_DATAFRAME:
            if len(tuple_list) == 1:
                # works for one column name
                column_name = self.data_frame.columns[tuple_list[0][1]]
                t = self.data_frame[column_name]

        else:

            if len(tuple_list) == 1:
                if tuple_list[0][1] == tuple_list[0][3]:
                    t = self.table[tuple_list[0][0]:tuple_list[0][2],
                                   tuple_list[0][3]]
                elif tuple_list[0][0] == tuple_list[0][2]:
                    t = self.table[tuple_list[0],
                                   tuple_list[0][1]:tuple_list[3]]
                t.reshape((1, t.size))
                for index in tuple_list[1:len(tuple_list) - 1]:
                    x = self.table[index[0]:index[2], index[1]:index[3]]
                    t = np.hstack((t, x.reshape((1, x.size))))

        self.selection_mean = t.mean()
        self.selection_std = t.std()
        self.selection_var = t.var()
        self.selection_sum = t.sum()
        self.selection_max = np.amax(t)
        self.selection_min = np.amin(t)

        self.selection_handler.flush()

    def create_pandas_dataframe(self):
        pass

    def string_ops(self):
        '''
        Perform trivial ops on a column of the data frame.
        '''
        self.selection_handler.create_selection()

        if len(self.selection_handler.selected_indices) == 1:
            column = self.selection_handler.selected_indices[0][1]
            # currently works only for a single column
            if self.AS_PANDAS_DATAFRAME:
                column_name = self.data_frame.columns[column]
                self.column_length = len(self.data_frame[column_name])
                self.unique_items_nos = self.data_frame[column_name].nunique()
            else:
                self.column_length = len(self.table[:, column])
                self.unique_items_nos = len(np.unique(self.table[:, column]))

        self.selection_handler.flush()

    def map_to_unique(self):
        '''
        USed to map a set of items onto unique integers,
        and change tabular display accordingly.
        '''
        self.selection_handler.create_selection()

        if len(self.selection_handler.selected_indices) == 1:
            # currently works only for a single coulmn
            self.unique_map = {}
            column = self.selection_handler.selected_indices[0][1]
            column_name = self.data_frame.columns[column]
            data = DataFrame(self.data_frame[column_name])
            m = 0

            gb = data.groupby([column_name])

            for key in gb.indices.keys():
                self.unique_map[key] = m
                m += 1
            new_indices = data[column_name]

            for i in range(len(new_indices)):
                key = new_indices[i]
                new_indices[i] = self.unique_map[key]
            self.data_frame[column_name] = new_indices

            self.table_model = DataFrameModel(
                self.data_frame,
                editable=True,
                horizontal_headers=self.data_frame.columns,
            )

        self.selection_handler.flush()

    def save_as(self):
        '''
        Save the current table as a csv file
        '''

        if self.AS_PANDAS_DATAFRAME:
            self.data_frame.to_csv(self.save_filename, index=False)

        else:
            # add file writing script for a numpy array
            pass

    def normalize_selection(self):
        '''
        Normalize the current selection.
        '''
        self.selection_handler.create_selection()
        if len(self.selection_handler.selected_indices) == 1:
            column_index = self.selection_handler.selected_indices[0][1]
            if self.AS_PANDAS_DATAFRAME:
                column_name = self.data_frame.columns[column_index]
                x = self.data_frame[column_name]
                x = x - x.mean()
                x = x / x.std()
                self.data_frame[column_name] = x
                self.table_model = DataFrameModel(
                    self.data_frame,
                    editable=True,
                    horizontal_headers=self.data_frame.columns)
            else:
                x = self.table[:, column_index]
                x = x - x.mean()
                x = x / x.std()
                self.table[:, column_index] = x
                self.table_model = TableModel(self.table,
                                              editable=True,
                                              horizontal_headers=self.headers)
        self.selection_handler.flush()

    def add_uservariables(self, var_dict):
        '''
        Used to add variables selected from the tabular view and named by
        the user to the local workspace. These variables are identifiable by
        their user given names in the script.
        '''

        if self.AS_PANDAS_DATAFRAME:
            for key in var_dict:

                top_left = var_dict[key][0]
                exec('top_left=' + top_left)

                bot_right = var_dict[key][1]
                exec('bot_right=' + bot_right)

                if top_left[1] == bot_right[1]:
                    # currently works only for columns
                    column = top_left[1]
                    col_name = self.data_frame.columns[column]
                    x = self.data_frame[col_name][top_left[0]:bot_right[0]]

                    self.script_handler.my_locals[key] = x

        else:
            for key in var_dict:

                top_left = var_dict[key][0]
                exec('top_left=' + top_left)

                bot_right = var_dict[key][1]
                exec('bot_right=' + bot_right)

                x = self.table[top_left[0]:bot_right[0] + 1,
                               top_left[1]:bot_right[1] + 1]

                self.script_handler.my_locals[key] = x

    def delete_rowcol(self):
        '''
        Called to delete columns from the pandas dataframe
        '''
        self.selection_handler.create_selection()

        to_remove = []
        for index in self.selection_handler.selected_indices:
            column_name = self.data_frame.columns[index[1]]
            to_remove.append(column_name)
        for column_name in to_remove:
            del self.data_frame[column_name]

        self.redraw_tablemodel()
        self.selection_handler.flush()

    def shift_selection(self, what, where):
        '''
        Called to shift a column to the left or to the right
        '''
        # currently works only for pandas dataframes
        if self.AS_PANDAS_DATAFRAME:
            if what == 'column':

                self.selection_handler.create_selection()
                column_list = list(self.data_frame.columns)
                selected_column_index = self.selection_handler.selected_indices[
                    0][1]
                selected_column = column_list.pop(selected_column_index)
                if where == 'left':
                    to_insert_index = selected_column_index - 1
                else:
                    to_insert_index = selected_column_index + 1
                column_list.insert(to_insert_index, selected_column)
                self.data_frame = self.data_frame.reindex(columns=column_list)

            elif what == 'row':

                self.selection_handler.create_selection()
                index = self.selection_handler.selected_indices[0][0]
                row_list = list(self.data_frame.index)
                selected_row = row_list.pop(index)
                if where == 'up':
                    to_insert_index = index - 1
                else:
                    to_insert_index = index + 1
                row_list.insert(to_insert_index, selected_row)
                self.data_frame = self.data_frame.reindex(index=row_list)

            self.redraw_tablemodel()

        self.selection_handler.flush()

    def sort_selection(self, sort_type):
        '''
        Called to sort the selected column
        '''
        # currently works only for pandas dataframes

        self.table_model.begin_change_layout()

        if self.AS_PANDAS_DATAFRAME:
            self.selection_handler.create_selection()
            column_index = self.selection_handler.selected_indices[0][1]
            column_name = self.data_frame.columns[column_index]
            column = self.data_frame[column_name].copy()
            if sort_type == 0:
                column = column.order()

            else:
                column = column.order(ascending=False)

            del self.data_frame[column_name]
            self.data_frame[column_name] = list(column)

        self.table_model.end_change_layout()

        self.redraw_tablemodel()

        self.selection_handler.flush()

    def add_row_col(self, selected_label):
        '''
        Adds a selected variable from the local workspace to the table view as
        a row or as a column.
        '''

        self.selection_handler.create_selection()
        selected_index = self.selection_handler.selected_indices[0][0]
        var_name = self.workspace_handler.workspace.keys()[selected_index]
        data = self.workspace_handler.workspace[var_name]

        if selected_label == 'As Row':
            if self.AS_PANDAS_DATAFRAME:
                if data.shape[1] == self.data_frame.shape[1]:
                    self.data_frame = self.data_frame.append(data)
                    print self.data_frame.shape
            else:
                if data.shape[1] == self.table.shape[1]:
                    self.table = np.vstack((self.table, data))
        else:

            if self.AS_PANDAS_DATAFRAME:
                if data.shape[0] == self.data_frame.shape[0]:
                    self.data_frame[var_name] = data
                    print self.data_frame.shape
            else:
                if data.shape[0] == self.table.shape[0]:
                    self.table = np.hstack((self.table, data))

        self.headers.append(var_name)
        self.selection_handler.flush()
        self.redraw_tablemodel()

    def set_table_selection(self, row, col):
        '''
        Sets a selection on the table view.
        '''
        none_types = [' ', 'None']

        if self.AS_PANDAS_DATAFRAME:
            data = self.data_frame
        else:
            data = self.table
        print row, col
        if (row in none_types) and (col not in none_types):
            top_left = (0, int(col))
            bot_right = (data.shape[0], int(col))
        elif (row not in none_types) and (col in none_types):
            top_left = (int(row), 0)
            bot_right = (int(row), data.shape[1])

        top_left_mi = self.table_model.create_index(row=top_left[0],
                                                    column=top_left[1],
                                                    context=None)
        bot_right_mi = self.table_model.create_index(row=bot_right[0],
                                                     column=bot_right[1],
                                                     context=None)

        return [(top_left_mi, bot_right_mi)]

    def increase_indent(self, selection):
        '''
        Increasing the indent by four spaces in the text editor.
        '''

        t = selection.split('\n')
        indented = ''
        for elem in t:
            elem = '    ' + elem + '\n'
            indented = indented + elem
        indented = indented.rstrip()
        return indented

    def decrease_indent(self, selection):
        pass

    def remove_workspace_selection(self):
        pass

    def save_workspace(self, file_path):
        '''
        Saves the local workspace as a .mat file. (All variables must be
        arrays or strings.)
        '''

        to_save = {}
        for elem in self.workspace_handler.workspace:
            obj = self.workspace_handler.workspace[elem]
            if isinstance(obj, np.ndarray):
                to_save[elem] = obj
        if len(to_save) > 0:
            savemat(file_path, to_save)

    def load_workspace(self, file_path):
        '''
        Loading a .mat file into the workspace.
        '''

        loaded_workspace = loadmat(file_path)
        for elem in loaded_workspace:
            self.workspace_handler.workspace[elem] = loaded_workspace[elem]
        self.script_handler.my_locals = self.workspace_handler.workspace

    def parse_timestamps(self, selected_option):
        '''
        Parse timestamps from the pandas dataframe into Pandas DatetimIndex
        objects or standard datetime objects.
        '''

        self.selection_handler.create_selection()
        column = self.selection_handler.selected_indices[0][1]
        column_name = self.data_frame.columns[column]
        data = self.data_frame[column_name]
        if selected_option == 1:
            index = DatetimeIndex(data)
            self.data_frame[column_name] = index
        self.selection_handler.flush()
Example #11
0
class CsvModel(HasTraits):
    '''
    The object that is passed to the MainWindow of the enaml view. This
    represents a general model for different plots in the csv editor.
    '''
    
    
    # The pandas data frame associated with the current instance of the editor.
    data_frame = Instance(DataFrame,())
    
    # Lenght of the selected column
    column_length = Int
    
    # Number of unique items in a selection
    unique_items_nos = Int
    
    # the unique mapping from a list of items to a list of integers
    unique_map = Dict

    # The .csv file to be opened.
    filename = File
    
    # The filename of the file to be saved
    save_filename = String
    
    # The numpy array associated with the data in the file
    table = Array
    
    # The headers of the csv data
    headers = List
    
    # The mean of the 2-D block of the array, represented by the current
    # selection
    selection_mean = Float
    
    # The variance of the 2-D block of the array, represented by the current
    # selection
    selection_var = Float
    
    # The standard deviation of the 2-D block of the array, represented by the
    # current selection
    selection_std = Float
    
    # The sum of the current selection
    selection_sum = Float
    
    # The minimum value from the current selection
    selection_min = Float
    
    # The maximum value from the current selection
    selection_max = Float
    
    # The covariance between a pair of rows or columns
    # NOT IMPLEMENTED
    selection_cov = Float(0)
    
    # The TableModel instance to be passed to the item_model attribute of the
    # TableView
    table_model = Instance(TableModel,())
    
    # class for handling scripts
    script_handler = Instance(ScriptHandler)
    
    # class for handling the workspace generated by the script
    workspace_handler = Any
    
    # Class for handling selections from the TableView
    selection_handler = Instance(SelectionHandler)
    
    # If the file should be imported as a Pandas dataframe
    AS_PANDAS_DATAFRAME = Bool(False)
    
    def _workspace_handler_default(self):
        return WorkspaceHandler({})
    
    def _script_handler_default(self):
        return ScriptHandler()
    
    def _selection_handler_default(self):
        return SelectionHandler()
    
    def _table_default(self):
        '''
        The default array in the TableView is a 100x100 array of zeros
        '''
        if not self.AS_PANDAS_DATAFRAME:
            x = np.zeros((100,100))
            return x
        else:
            return None
    
    def _data_frame_default(self):
        if self.AS_PANDAS_DATAFRAME:
            return DataFrame()
        else:
            return None
    
    def _table_model_default(self):
        '''
        The default TableModel instance corresponding to the array of zeros
        '''
        
        tblmodel =TableModel(self.table, editable=True)
        return tblmodel
    
    def _image_plot_default(self):
        '''
        Default chaco plot object for the image plot.
        '''
        
        self.img_plotdata = ArrayPlotData(imagedata=self.table)
        p = Plot(self.img_plotdata)
        p.img_plot('imagedata')
        return p
    
    def _hist_plot_default(self):
        '''
        Default chaco plot object for the histogram.
        '''
        
        h = np.histogram(self.table,10)[0]
        self.hist_plotdata = ArrayPlotData(x=h)
        p = Plot(self.hist_plotdata)
        p.plot('x',type='bar',color='auto',bar_width=0.3)
        return p
    
    def _filename_changed(self, new):
        '''
        Executes whenever a file is loaded into the view.
        '''
        if self.AS_PANDAS_DATAFRAME:
            self.data_frame = read_csv(self.filename)
            self.table_model = DataFrameModel(
                self.data_frame,
                editable=True,
                horizontal_headers=self.data_frame.columns,
            )
        else:
            self.table = np.genfromtxt(self.filename, delimiter=',', skip_header=1)
            csv_reader= csv.reader(file(self.filename))
            self.headers = csv_reader.next()
            del csv_reader
            self.table_model = TableModel(self.table, editable=True,
                                     horizontal_headers=self.headers)
    
    
    def redraw_tablemodel(self):
        '''
        Called when the tabular view needs to be redrawn when the changes to the
        table are programmatic, not through the UI.
        '''
        
        if self.AS_PANDAS_DATAFRAME:
            self.table_model = DataFrameModel(
                self.data_frame, editable= True,
                horizontal_headers=self.data_frame.columns
            )
        else:
            self.table_model = TableModel(self.table, editable=True,
                                          horizontal_headers=self.headers)
    
    def _pca_default(self):
        return PCA(2, whiten=True)
    
    def create_plot_properties():
        pass
    
    
    def use_selection_preview(self):
        pass
    
    
    def use_selection_histogram(self):
        pass
    
    def use_selection_imageplot(self):
        pass
    
    def use_selection_pcaplot(self):
        pass
    
    def calculate_selection_params(self):
        
        '''
        Called to calculate statistics of a selection.
        '''
        self.selection_handler.create_selection()
        tuple_list = self.selection_handler.selected_indices

        if self.AS_PANDAS_DATAFRAME:
            if len(tuple_list)==1:
                # works for one column name
                column_name = self.data_frame.columns[tuple_list[0][1]]
                t = self.data_frame[column_name]
                
                
        else:
            
            if len(tuple_list) == 1:
                if tuple_list[0][1]==tuple_list[0][3]:
                    t = self.table[tuple_list[0][0]:tuple_list[0][2],
                                   tuple_list[0][3]]
                elif tuple_list[0][0]==tuple_list[0][2]:
                    t = self.table[tuple_list[0],
                                   tuple_list[0][1]:tuple_list[3]]
                t.reshape((1,t.size))
                for index in tuple_list[1:len(tuple_list)-1]:
                    x = self.table[index[0]:index[2],index[1]:index[3]]
                    t = np.hstack((t,x.reshape((1,x.size))))

        self.selection_mean = t.mean()
        self.selection_std = t.std()
        self.selection_var = t.var()
        self.selection_sum = t.sum()
        self.selection_max = np.amax(t)
        self.selection_min = np.amin(t)
        
        
        self.selection_handler.flush()
    
    def create_pandas_dataframe(self):
        pass
    
    def string_ops(self):
        '''
        Perform trivial ops on a column of the data frame.
        '''
        self.selection_handler.create_selection()
        
        if len(self.selection_handler.selected_indices)==1:
            column = self.selection_handler.selected_indices[0][1]
            # currently works only for a single column
            if self.AS_PANDAS_DATAFRAME:                
                column_name = self.data_frame.columns[column]
                self.column_length = len(self.data_frame[column_name])
                self.unique_items_nos = self.data_frame[column_name].nunique()
            else:
                self.column_length = len(self.table[:,column])
                self.unique_items_nos = len(np.unique(self.table[:,column]))
        
        self.selection_handler.flush()
    
    def map_to_unique(self):
        '''
        USed to map a set of items onto unique integers,
        and change tabular display accordingly.
        '''
        self.selection_handler.create_selection()
        
        if len(self.selection_handler.selected_indices)==1:
            # currently works only for a single coulmn
            self.unique_map = {}
            column = self.selection_handler.selected_indices[0][1]
            column_name = self.data_frame.columns[column]
            data = DataFrame(self.data_frame[column_name])
            m = 0
            
            gb = data.groupby([column_name])
            
            for key in gb.indices.keys():
                self.unique_map[key] = m
                m += 1
            new_indices = data[column_name]
            
            for i in range(len(new_indices)):
                key = new_indices[i]
                new_indices[i] = self.unique_map[key]
            self.data_frame[column_name] = new_indices
            
            self.table_model = DataFrameModel(
                self.data_frame,
                editable=True,
                horizontal_headers=self.data_frame.columns,
            )
        
        self.selection_handler.flush()
    
    def save_as(self):
        '''
        Save the current table as a csv file
        '''
        
        if self.AS_PANDAS_DATAFRAME:
            self.data_frame.to_csv(self.save_filename,index=False)
            
        else:
            # add file writing script for a numpy array
            pass
        
    
    def normalize_selection(self):
        '''
        Normalize the current selection.
        '''
        self.selection_handler.create_selection()
        if len(self.selection_handler.selected_indices)==1:
            column_index = self.selection_handler.selected_indices[0][1]
            if self.AS_PANDAS_DATAFRAME:
                column_name = self.data_frame.columns[column_index]
                x = self.data_frame[column_name]
                x = x - x.mean()
                x = x/x.std()
                self.data_frame[column_name]=x
                self.table_model = DataFrameModel(
                    self.data_frame,
                    editable=True,
                    horizontal_headers=self.data_frame.columns
                )
            else:
                x = self.table[:,column_index]
                x = x - x.mean()
                x = x/x.std()
                self.table[:,column_index] = x
                self.table_model = TableModel(self.table, editable=True,
                                              horizontal_headers=self.headers)
        self.selection_handler.flush()

    
    def add_uservariables(self, var_dict):
        '''
        Used to add variables selected from the tabular view and named by
        the user to the local workspace. These variables are identifiable by
        their user given names in the script.
        '''
        
        if self.AS_PANDAS_DATAFRAME:
            for key in var_dict:
                
                top_left = var_dict[key][0]
                exec('top_left='+top_left)
                
                
                bot_right = var_dict[key][1]
                exec('bot_right='+bot_right)
                
                if top_left[1] == bot_right[1]:
                    # currently works only for columns
                    column = top_left[1]
                    col_name = self.data_frame.columns[column]
                    x = self.data_frame[col_name][top_left[0]:bot_right[0]]
                    
                    self.script_handler.my_locals[key] = x
                    
        
        else:
            for key in var_dict:
                
                top_left = var_dict[key][0]
                exec('top_left='+top_left)
                
                
                bot_right = var_dict[key][1]
                exec('bot_right='+bot_right)
                
                
                x = self.table[top_left[0]:bot_right[0]+1,
                               top_left[1]:bot_right[1]+1]
                
                self.script_handler.my_locals[key] = x
            
    
    def delete_rowcol(self):
        '''
        Called to delete columns from the pandas dataframe
        '''
        self.selection_handler.create_selection()
        
        to_remove = []        
        for index in self.selection_handler.selected_indices:
            column_name = self.data_frame.columns[index[1]]
            to_remove.append(column_name)
        for column_name in to_remove:
            del self.data_frame[column_name]
        
        self.redraw_tablemodel()
        self.selection_handler.flush()
    
    def shift_selection(self, what, where):
        '''
        Called to shift a column to the left or to the right
        '''
        # currently works only for pandas dataframes
        if self.AS_PANDAS_DATAFRAME:
            if what == 'column':
            
                self.selection_handler.create_selection()
                column_list = list(self.data_frame.columns)
                selected_column_index = self.selection_handler.selected_indices[0][1]
                selected_column = column_list.pop(selected_column_index)
                if where == 'left':
                    to_insert_index = selected_column_index - 1
                else:
                    to_insert_index = selected_column_index + 1
                column_list.insert(to_insert_index, selected_column)
                self.data_frame = self.data_frame.reindex(columns = column_list)
        
        
            elif what == 'row':
        
                self.selection_handler.create_selection()
                index = self.selection_handler.selected_indices[0][0]
                row_list = list(self.data_frame.index)
                selected_row = row_list.pop(index)
                if where == 'up':
                    to_insert_index = index - 1
                else:
                    to_insert_index = index + 1
                row_list.insert(to_insert_index, selected_row)
                self.data_frame = self.data_frame.reindex(index=row_list)
        
            self.redraw_tablemodel()        
        
        self.selection_handler.flush()
    
    def sort_selection(self, sort_type):
        '''
        Called to sort the selected column
        '''
        # currently works only for pandas dataframes
        
        self.table_model.begin_change_layout()
        
        if self.AS_PANDAS_DATAFRAME:
            self.selection_handler.create_selection()
            column_index = self.selection_handler.selected_indices[0][1]
            column_name = self.data_frame.columns[column_index]
            column = self.data_frame[column_name].copy()
            if sort_type == 0:
                column = column.order()
                
            else:
                column = column.order(ascending=False)
            
            del self.data_frame[column_name]
            self.data_frame[column_name] = list(column)
                
            
        self.table_model.end_change_layout()
        
        self.redraw_tablemodel()
        
        
        
        self.selection_handler.flush()
    
    def add_row_col(self, selected_label):
        '''
        Adds a selected variable from the local workspace to the table view as
        a row or as a column.
        '''
        
        self.selection_handler.create_selection()
        selected_index = self.selection_handler.selected_indices[0][0]
        var_name = self.workspace_handler.workspace.keys()[selected_index]
        data = self.workspace_handler.workspace[var_name]

        if selected_label == 'As Row':
            if self.AS_PANDAS_DATAFRAME:
                if data.shape[1] == self.data_frame.shape[1]:
                    self.data_frame = self.data_frame.append(data)
                    print self.data_frame.shape
            else:
                if data.shape[1] == self.table.shape[1]:
                    self.table = np.vstack((self.table, data))
        else :
            
            if self.AS_PANDAS_DATAFRAME:
                if data.shape[0] == self.data_frame.shape[0]:
                    self.data_frame[var_name] = data
                    print self.data_frame.shape
            else:
                if data.shape[0] == self.table.shape[0]:
                    self.table = np.hstack((self.table, data))
        
        self.headers.append(var_name)        
        self.selection_handler.flush()
        self.redraw_tablemodel()
        
    
    def set_table_selection(self, row, col):
        '''
        Sets a selection on the table view.
        '''
        none_types = [' ', 'None']
        
        
        
        if self.AS_PANDAS_DATAFRAME:
            data = self.data_frame
        else:
            data = self.table
        print row, col
        if (row in none_types) and (col not in none_types):
            top_left = (0, int(col))
            bot_right = (data.shape[0], int(col))
        elif (row not in none_types) and (col in none_types):
            top_left = (int(row), 0)
            bot_right = (int(row), data.shape[1])
        
        top_left_mi = self.table_model.create_index(row=top_left[0],
                                                    column=top_left[1],
                                                    context=None)
        bot_right_mi = self.table_model.create_index(row=bot_right[0],
                                                    column=bot_right[1],
                                                    context=None)
        
        return [(top_left_mi, bot_right_mi)]
    
    def increase_indent(self, selection):
        '''
        Increasing the indent by four spaces in the text editor.
        '''
        
        t = selection.split('\n')
        indented = ''
        for elem in t:
            elem = '    '+elem+'\n'
            indented = indented + elem
        indented = indented.rstrip()
        return indented
    
    def decrease_indent(self, selection):
        pass
    
    def remove_workspace_selection(self):
        pass
    
    def save_workspace(self, file_path):
        '''
        Saves the local workspace as a .mat file. (All variables must be
        arrays or strings.)
        '''
        
        to_save = {}
        for elem in self.workspace_handler.workspace:
            obj = self.workspace_handler.workspace[elem]
            if isinstance(obj, np.ndarray):
                to_save[elem] = obj
        if len(to_save)>0:
            savemat(file_path, to_save)
    
    def load_workspace(self, file_path):
        '''
        Loading a .mat file into the workspace.
        '''
        
        loaded_workspace = loadmat(file_path)
        for elem in loaded_workspace:
            self.workspace_handler.workspace[elem] = loaded_workspace[elem]
        self.script_handler.my_locals = self.workspace_handler.workspace

    def parse_timestamps(self, selected_option):
        '''
        Parse timestamps from the pandas dataframe into Pandas DatetimIndex
        objects or standard datetime objects.
        '''
        
        self.selection_handler.create_selection()
        column = self.selection_handler.selected_indices[0][1]
        column_name = self.data_frame.columns[column]
        data = self.data_frame[column_name]
        if selected_option == 1:
            index = DatetimeIndex(data)
            self.data_frame[column_name] = index
        self.selection_handler.flush()