Пример #1
0
def analysis_pdf(datafile, Nmax=1000000):

    pdf_filename = os.path.join(summary_pdf_folder(datafile), 'rois.pdf')

    data = MultimodalData(datafile)
    data.correctedFluo, data.F0 = data.build_dFoF(
        return_corrected_F_and_F0=True)

    with PdfPages(pdf_filename) as pdf:

        print('* plotting ROI analysis as "rois.pdf" [...]')
        print('   - plotting imaging FOV')
        fig, AX = plt.subplots(1, 5, figsize=(11.4, 2.5))
        plt.subplots_adjust(left=0.03, right=0.97)
        data.show_CaImaging_FOV(key='meanImg', NL=1, cmap='viridis', ax=AX[0])
        data.show_CaImaging_FOV(key='meanImg', NL=2, cmap='viridis', ax=AX[1])
        data.show_CaImaging_FOV(key='meanImgE', NL=2, cmap='viridis', ax=AX[2])
        data.show_CaImaging_FOV(key='max_proj', NL=2, cmap='viridis', ax=AX[3])
        try:
            data.show_CaImaging_FOV(key='meanImg_chan2',
                                    NL=2,
                                    cmap='viridis',
                                    ax=AX[4])
        except KeyError:
            AX[4].annotate('no red channel', (0.5, 0.5),
                           xycoords='axes fraction',
                           fontsize=9,
                           va='center',
                           ha='center')
            AX[4].axis('off')
        pdf.savefig()  # saves the current figure into a pdf page
        plt.close()
        for i in np.arange(data.nROIs)[:Nmax]:
            print('   - plotting analysis of ROI #%i' % (i + 1))
            try:
                fig = raw_fluo_fig(data, roiIndex=i)
                pdf.savefig(fig)  # saves the current figure into a pdf page
                plt.close(fig)
                fig = analysis_fig(data, roiIndex=i)
                pdf.savefig(fig)  # saves the current figure into a pdf page
                plt.close(fig)
            except BaseException as be:
                print(be)
                print('  /!\ Pb with ROI #%i /!\ ' % (i + 1))

    print('[ok] roi analysis saved as: "%s" ' % pdf_filename)
Пример #2
0
class FiguresWindow(NewWindow):
    def __init__(
            self,
            parent=None,
            dt_sampling=10,  # ms
            fig_name=os.path.join(os.path.expanduser('~'), 'Desktop',
                                  'fig.svg'),
            title='Figures'):

        super(FiguresWindow, self).__init__(parent=parent.app,
                                            title=title,
                                            i=-10,
                                            size=(800, 800))

        self.modalities = []
        for key1, key2 in zip([
                'Photodiode-Signal', 'Electrophysiological-Signal',
                'Running-Speed', 'Pupil', 'CaImaging-TimeSeries',
                'CaImaging-TimeSeries', 'Photodiode-Signal'
        ], [
                'Photodiode', 'Electrophy', 'Locomotion', 'Pupil',
                'CaImagingSum', 'CaImaging', 'VisualStim'
        ]):
            if key1 in parent.nwbfile.acquisition:
                self.modalities.append(key2)

        self.parent = parent
        self.get_varied_parameters()
        self.fig_name = fig_name

        self.data = MultimodalData(self.parent.datafile)

        self.EPISODES = None
        self.xlim, self.ylim = [-10, 10], [-10, 10]

        mainLayout, Layouts = QtWidgets.QVBoxLayout(self.cwidget), []

        # -- protocol
        Layouts.append(QtWidgets.QHBoxLayout())
        # Layouts[-1].addWidget(QtWidgets.QLabel('-- Protocol:   ', self))
        self.pbox = QtWidgets.QComboBox(self)
        self.pbox.setFixedWidth(400)
        self.pbox.addItem(' [PROTOCOL] ')
        self.pbox.addItems(self.parent.protocols)
        Layouts[-1].addWidget(self.pbox)

        # -- quantity and subquantity
        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(QtWidgets.QLabel('-- Quantity:   ', self))
        self.qbox = QtWidgets.QComboBox(self)
        self.qbox.addItem('')
        if 'ophys' in self.parent.nwbfile.processing:
            self.qbox.addItem('CaImaging')
        for key in parent.nwbfile.acquisition:
            if len(parent.nwbfile.acquisition[key].data.shape) == 1:
                self.qbox.addItem(key)  # only for scalar variables
        self.qbox.setFixedWidth(400)
        Layouts[-1].addWidget(self.qbox)
        Layouts[-1].addWidget(
            QtWidgets.QLabel('       -- Sub-quantity:   ', self))
        self.sqbox = QtWidgets.QLineEdit(self)
        self.sqbox.setText(35 * ' ' + 'e.g. "dF/F", "Neuropil", "pLFP", ...')
        self.sqbox.setMinimumWidth(400)
        Layouts[-1].addWidget(self.sqbox)

        # -- roi
        if 'ophys' in self.parent.nwbfile.processing:
            Layouts[-1].addWidget(QtWidgets.QLabel('    -- ROI:   ', self))
            self.roiPick = QtGui.QLineEdit()
            self.roiPick.setText(
                '    [select ROI]  e.g.: "1",  "10-20", "3, 45, 7", ... ')
            self.roiPick.setMinimumWidth(400)
            self.roiPick.returnPressed.connect(self.select_ROI)
            Layouts[-1].addWidget(self.roiPick)

        ### PLOT RAW DATA
        Layouts.append(QtWidgets.QHBoxLayout())
        self.rawPlotBtn = QtWidgets.QPushButton(' plot raw data ', self)
        self.rawPlotBtn.setFixedWidth(200)
        self.rawPlotBtn.clicked.connect(self.plot_raw_data)
        Layouts[-1].addWidget(self.rawPlotBtn)
        for mod in self.modalities:
            Layouts[-1].addWidget(QtWidgets.QLabel(' ' + mod + ':', self))
            setattr(self, mod + 'Plot', QtWidgets.QDoubleSpinBox(self))
            getattr(self, mod + 'Plot').setValue(1)
            getattr(self, mod + 'Plot').setSuffix(' (%-fig)')
            Layouts[-1].addWidget(getattr(self, mod + 'Plot'))

        ### PLOT FIELD OF VIEW
        if 'ophys' in self.parent.nwbfile.processing:
            Layouts.append(QtWidgets.QHBoxLayout())
            self.fovPlotBtn = QtWidgets.QPushButton(
                'plot imaging field of view ', self)
            self.fovPlotBtn.setFixedWidth(250)
            self.fovPlotBtn.clicked.connect(self.plot_FOV)
            Layouts[-1].addWidget(self.fovPlotBtn)
            Layouts[-1].addWidget(
                QtWidgets.QLabel(10 * ' ' + 'FOV type:', self))  # SEPARATOR
            self.fovType = QtWidgets.QComboBox(self)
            self.fovType.addItems(['meanImg', 'meanImgE', 'max_proj'])
            Layouts[-1].addWidget(self.fovType)
            Layouts[-1].addWidget(QtWidgets.QLabel(10 * ' ' + 'NL:',
                                                   self))  # SEPARATOR
            self.fovNL = QtWidgets.QSpinBox(self)
            self.fovNL.setValue(1)
            Layouts[-1].addWidget(self.fovNL)
            Layouts[-1].addWidget(
                QtWidgets.QLabel(10 * ' ' + 'colormap:', self))  # SEPARATOR
            self.fovMap = QtWidgets.QComboBox(self)
            self.fovMap.addItems(
                ['viridis', 'grey', 'white_to_green', 'black_to_green'])
            Layouts[-1].addWidget(self.fovMap)
            Layouts[-1].addWidget(QtWidgets.QLabel('  label:',
                                                   self))  # SEPARATOR
            self.fovLabel = QtGui.QLineEdit()
            Layouts[-1].addWidget(self.fovLabel)

        ### PLOT CORRELATIONS
        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(QtWidgets.QLabel(50 * '<->', self))  # SEPARATOR
        Layouts[-1].addWidget(QtWidgets.QLabel('  -* CORRELATIONS *-', self))

        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(QtWidgets.QLabel(50 * '<->', self))  # SEPARATOR

        # varied keys
        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(
            QtWidgets.QLabel('  -* STIMULI PARAMETERS *-', self))
        for key in self.varied_parameters:
            Layouts.append(QtWidgets.QHBoxLayout())
            Layouts[-1].addWidget(QtWidgets.QLabel('--- %s ' % key, self))
            setattr(self, '%s_plot' % key, QtWidgets.QComboBox(self))
            getattr(self, '%s_plot' % key).addItems([
                'merged', 'single-value', 'column-panels', 'row-panels',
                'color-coded', 'x-axis', 'N/A'
            ])
            Layouts[-1].addWidget(getattr(self, '%s_plot' % key))
            setattr(self, '%s_values' % key, QtWidgets.QComboBox(self))
            getattr(
                self, '%s_values' %
                key).addItems(['full', 'custom'] +
                              [str(s) for s in self.varied_parameters[key]])
            Layouts[-1].addWidget(getattr(self, '%s_values' % key))
            Layouts[-1].addWidget(QtWidgets.QLabel(10 * ' ', self))
            Layouts[-1].addWidget(QtWidgets.QLabel(' custom values : ', self))
            setattr(self, '%s_customvalues' % key, QtWidgets.QLineEdit(self))
            getattr(self, '%s_customvalues' % key).setMaximumWidth(150)
            Layouts[-1].addWidget(getattr(self, '%s_customvalues' % key))
            Layouts[-1].addWidget(QtWidgets.QLabel(30 * ' ', self))

        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(QtWidgets.QLabel(50 * '<->', self))  # SEPARATOR

        # figure props type
        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(
            QtWidgets.QLabel('       -* RESPONSES *-        ', self))
        self.responseType = QtWidgets.QComboBox(self)
        self.responseType.addItems(
            ['stim-evoked-traces', 'mean-stim-evoked', 'integral-stim-evoked'])
        Layouts[-1].addWidget(self.responseType)
        Layouts[-1].addWidget(QtWidgets.QLabel('       color:  ', self))
        self.color = QtGui.QLineEdit()
        Layouts[-1].addWidget(self.color)
        Layouts[-1].addWidget(QtWidgets.QLabel('       label:  ', self))
        self.label = QtGui.QLineEdit()
        Layouts[-1].addWidget(self.label)
        Layouts[-1].addWidget(QtWidgets.QLabel('    n-label:  ', self))
        self.nlabel = QtGui.QLineEdit()
        self.nlabel.setText('1')
        Layouts[-1].addWidget(self.nlabel)
        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(
            QtWidgets.QLabel(' pre-stimulus window (s):', self))
        self.preWindow = QtGui.QLineEdit()
        self.preWindow.setText('[-1,0]')
        Layouts[-1].addWidget(self.preWindow)
        Layouts[-1].addWidget(
            QtWidgets.QLabel(' post-stimulus window (s):', self))
        self.postWindow = QtGui.QLineEdit()
        self.postWindow.setText('[1,4]')
        Layouts[-1].addWidget(self.postWindow)
        Layouts.append(QtWidgets.QHBoxLayout())
        self.baseline = QtGui.QCheckBox("baseline substraction")
        Layouts[-1].addWidget(self.baseline)
        Layouts[-1].addWidget(QtWidgets.QLabel(10 * '  ', self))
        self.withStatTest = QtGui.QCheckBox("with stat. test")
        Layouts[-1].addWidget(self.withStatTest)

        Layouts.append(QtWidgets.QHBoxLayout())
        self.compute = QtWidgets.QPushButton('compute episodes', self)
        self.compute.setFixedWidth(200)
        self.compute.clicked.connect(self.compute_episodes)
        Layouts[-1].addWidget(self.compute)

        self.plotBtn = QtWidgets.QPushButton(' -* PLOT *- ', self)
        self.plotBtn.setFixedWidth(200)
        self.plotBtn.clicked.connect(self.plot)
        Layouts[-1].addWidget(self.plotBtn)

        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(QtWidgets.QLabel(50 * '<->', self))  # SEPARATOR

        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(
            QtWidgets.QLabel('       -* Figure Properties *-        ', self))
        self.fig_presets = QtWidgets.QComboBox(self)
        self.fig_presets.setFixedWidth(400)
        self.fig_presets.addItems(
            ['', 'raw-traces-preset', 'IO-curves-preset'])
        Layouts[-1].addWidget(self.fig_presets)

        Layouts[-1].addWidget(QtWidgets.QLabel('Panel size:  ', self))
        self.panelsize = QtGui.QLineEdit()
        self.panelsize.setText('(1,1)')
        Layouts[-1].addWidget(self.panelsize)

        #
        self.samplingBox = QtWidgets.QDoubleSpinBox(self)
        self.samplingBox.setValue(dt_sampling)
        self.samplingBox.setMaximum(500)
        self.samplingBox.setMinimum(0.1)
        self.samplingBox.setSuffix(' (ms) sampling')
        Layouts[-1].addWidget(self.samplingBox)

        # plot type
        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(QtWidgets.QLabel('Plot type: ', self))
        self.plotbox = QtWidgets.QComboBox(self)
        self.plotbox.addItems(['2d-plot', 'polar-plot', 'bar-plot'])
        Layouts[-1].addWidget(self.plotbox)
        Layouts[-1].addWidget(QtWidgets.QLabel(10 * '  ', self))
        self.plotbox2 = QtWidgets.QComboBox(self)
        self.plotbox2.addItems(['line', 'dot'])
        Layouts[-1].addWidget(self.plotbox2)
        Layouts[-1].addWidget(QtWidgets.QLabel(10 * '  ', self))
        self.withSTDbox = QtGui.QCheckBox("with s.d.")
        Layouts[-1].addWidget(self.withSTDbox)
        Layouts[-1].addWidget(QtWidgets.QLabel(10 * '  ', self))
        self.stim = QtGui.QCheckBox("with stim. ")
        Layouts[-1].addWidget(self.stim)
        Layouts[-1].addWidget(QtWidgets.QLabel(10 * '  ', self))
        self.screen = QtGui.QCheckBox("with screen inset ")
        Layouts[-1].addWidget(self.screen)
        Layouts[-1].addWidget(QtWidgets.QLabel(10 * '  ', self))
        self.annot = QtGui.QCheckBox("with annot. ")
        Layouts[-1].addWidget(self.annot)
        Layouts[-1].addWidget(QtWidgets.QLabel(10 * '  ', self))
        self.axis = QtGui.QCheckBox("with axis")
        Layouts[-1].addWidget(self.axis)
        Layouts[-1].addWidget(QtWidgets.QLabel(10 * '  ', self))
        self.grid = QtGui.QCheckBox("with grid")
        Layouts[-1].addWidget(self.grid)

        # X-scales
        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(QtWidgets.QLabel('X-SCALE      ---     ', self))
        Layouts[-1].addWidget(QtWidgets.QLabel('x-min:', self))
        self.xmin = QtGui.QLineEdit()
        self.xmin.setText('')
        Layouts[-1].addWidget(self.xmin)
        Layouts[-1].addWidget(QtWidgets.QLabel('x-max:', self))
        self.xmax = QtGui.QLineEdit()
        self.xmax.setText('')
        Layouts[-1].addWidget(self.xmax)
        Layouts[-1].addWidget(QtWidgets.QLabel('x-bar:', self))
        self.xbar = QtGui.QLineEdit()
        self.xbar.setText('')
        Layouts[-1].addWidget(self.xbar)
        Layouts[-1].addWidget(QtWidgets.QLabel('x-barlabel:', self))
        self.xbarlabel = QtGui.QLineEdit()
        self.xbarlabel.setText('')
        Layouts[-1].addWidget(self.xbarlabel)

        # Y-scales
        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(QtWidgets.QLabel('Y-SCALE      ---     ', self))
        Layouts[-1].addWidget(QtWidgets.QLabel('y-min:', self))
        self.ymin = QtGui.QLineEdit()
        self.ymin.setText('')
        Layouts[-1].addWidget(self.ymin)
        Layouts[-1].addWidget(QtWidgets.QLabel('y-max:', self))
        self.ymax = QtGui.QLineEdit()
        self.ymax.setText('')
        Layouts[-1].addWidget(self.ymax)
        Layouts[-1].addWidget(QtWidgets.QLabel('y-bar:', self))
        self.ybar = QtGui.QLineEdit()
        self.ybar.setText('')
        Layouts[-1].addWidget(self.ybar)
        Layouts[-1].addWidget(QtWidgets.QLabel('y-barlabel:', self))
        self.ybarlabel = QtGui.QLineEdit()
        self.ybarlabel.setText('')
        Layouts[-1].addWidget(self.ybarlabel)

        # curve
        Layouts.append(QtWidgets.QHBoxLayout())
        # self.Layout12.addWidget(self.baselineCB)

        Layouts.append(QtWidgets.QHBoxLayout())
        Layouts[-1].addWidget(QtWidgets.QLabel(50 * '<->', self))  # SEPARATOR

        # BUTTONS
        Layouts.append(QtWidgets.QHBoxLayout())
        self.setBtn = QtWidgets.QPushButton('extract settings', self)
        self.setBtn.setFixedWidth(200)
        self.setBtn.clicked.connect(self.set_settings)
        Layouts[-1].addWidget(self.setBtn)
        self.rstBtn = QtWidgets.QPushButton('reset settings', self)
        self.rstBtn.setFixedWidth(200)
        self.rstBtn.clicked.connect(self.reset_settings)
        Layouts[-1].addWidget(self.rstBtn)

        Layouts.append(QtWidgets.QHBoxLayout())
        self.newFig = QtWidgets.QPushButton('save as new figure', self)
        self.newFig.setFixedWidth(200)
        self.newFig.clicked.connect(self.new_fig)
        Layouts[-1].addWidget(self.newFig)

        self.append2Fig = QtWidgets.QPushButton('append to figure', self)
        self.append2Fig.setFixedWidth(200)
        self.append2Fig.clicked.connect(self.append)
        Layouts[-1].addWidget(self.append2Fig)

        self.exportBtn = QtWidgets.QPushButton('export to png', self)
        self.exportBtn.setFixedWidth(200)
        self.exportBtn.clicked.connect(self.export)
        Layouts[-1].addWidget(self.exportBtn)

        self.locBtn = QtWidgets.QComboBox(self)
        self.locBtn.addItems(['Desktop', 'summary'])
        self.locBtn.setFixedWidth(200)
        Layouts[-1].addWidget(self.locBtn)

        self.nameBtn = QtWidgets.QLineEdit(self)
        self.nameBtn.setText('fig')
        self.nameBtn.setFixedWidth(200)
        Layouts[-1].addWidget(self.nameBtn)

        self.dpi = QtWidgets.QSpinBox(self)
        self.dpi.setValue(300)
        self.dpi.setRange(50, 600)
        self.dpi.setSuffix(' (dpi)')
        self.dpi.setFixedWidth(80)
        Layouts[-1].addWidget(self.dpi)

        for l in Layouts:
            mainLayout.addLayout(l)

    def set_fig_name(self):
        if self.locBtn.currentText() == 'Desktop':
            self.fig_name = os.path.join(os.path.expanduser('~'), 'Desktop',
                                         self.nameBtn.text() + '.svg')
        elif self.locBtn.currentText() == 'summary':
            summary_dir = os.path.join(
                os.path.dirname(self.parent.datafile), 'summary',
                os.path.basename(self.parent.datafile).replace('.nwb', ''))
            pathlib.Path(summary_dir).mkdir(parents=True, exist_ok=True)
            self.fig_name = os.path.join(summary_dir,
                                         self.nameBtn.text() + '.svg')

    def export(self):
        export_drawing_as_png(self.fig_name,
                              dpi=self.dpi.value(),
                              background='white')

    def append(self):
        add_plot_to_svg(self.fig, self.fig_name)

    def new_fig(self):
        self.set_fig_name()
        self.fig.savefig(self.fig_name, transparent=True)

    def get_varied_parameters(self):
        self.varied_parameters = {}
        for key in self.parent.nwbfile.stimulus.keys():
            if key not in [
                    'frame_run_type', 'index', 'protocol_id', 'time_duration',
                    'time_start', 'time_start_realigned', 'time_stop',
                    'time_stop_realigned', 'interstim', 'interstim-screen'
            ]:
                unique = np.unique(self.parent.nwbfile.stimulus[key].data[:])
                if len(unique) > 1:
                    self.varied_parameters[key] = unique
        self.show()

    def set_settings(self):
        self.ymin.setText(str(self.ylim[0]))
        self.ymax.setText(str(self.ylim[1]))
        self.xmin.setText(str(self.xlim[0]))
        self.xmax.setText(str(self.xlim[1]))
        dx, dy = 0.1 * (self.xlim[1] - self.xlim[0]), 0.2 * (self.xlim[1] -
                                                             self.xlim[0])
        self.xbar.setText('%.1f' % dx)
        self.xbarlabel.setText('%.1f' % dx)
        self.ybar.setText('%.1f' % dy)
        self.ybarlabel.setText('%.1f' % dy)

    def reset_settings(self):
        for x in [
                self.xmin, self.xmax, self.ymin, self.ymax, self.xbar,
                self.ybar, self.xbarlabel, self.ybarlabel
        ]:
            x.setText('')

    def plot_row_column_of_quantity(self):

        single_cond = self.build_single_conditions()
        COL_CONDS = self.build_column_conditions()
        ROW_CONDS = self.build_row_conditions()
        COLOR_CONDS = self.build_color_conditions()

        if (len(COLOR_CONDS) > 1) and (self.color.text() != ''):
            COLORS = [
                getattr(ge, self.color.text())((c % 10) / 10.)
                for c in np.arange(len(COLOR_CONDS))
            ]
        elif (len(COLOR_CONDS) > 1):
            COLORS = [
                ge.tab10((c % 10) / 10.) for c in np.arange(len(COLOR_CONDS))
            ]
        elif self.color.text() != '':
            COLORS = [getattr(ge, self.color.text())]
        else:
            COLORS = ['k']

        if self.fig_presets.currentText() == 'raw-traces-preset':
            fig, AX = ge.figure(axes=(len(COL_CONDS), len(ROW_CONDS)),
                                reshape_axes=False,
                                top=0.4,
                                bottom=0.4,
                                left=0.7,
                                right=0.7,
                                wspace=0.5,
                                hspace=0.5)
        else:
            fig, AX = ge.figure(axes=(len(COL_CONDS), len(ROW_CONDS)),
                                reshape_axes=False)

        self.ylim = [np.inf, -np.inf]
        for irow, row_cond in enumerate(ROW_CONDS):
            for icol, col_cond in enumerate(COL_CONDS):
                for icolor, color_cond in enumerate(COLOR_CONDS):
                    cond = np.array(
                        single_cond & col_cond & row_cond
                        & color_cond)[:self.EPISODES['resp'].shape[0]]
                    if self.EPISODES['resp'][cond, :].shape[0] > 0:
                        my = self.EPISODES['resp'][cond, :].mean(axis=0)
                        if self.withSTDbox.isChecked():
                            sy = self.EPISODES['resp'][cond, :].std(axis=0)
                            ge.plot(self.EPISODES['t'],
                                    my,
                                    sy=sy,
                                    ax=AX[irow][icol],
                                    color=COLORS[icolor],
                                    lw=1)
                            self.ylim = [
                                min([self.ylim[0],
                                     np.min(my - sy)]),
                                max([self.ylim[1],
                                     np.max(my + sy)])
                            ]
                        else:
                            AX[irow][icol].plot(self.EPISODES['t'],
                                                my,
                                                color=COLORS[icolor],
                                                lw=1)
                            self.ylim = [
                                min([self.ylim[0], np.min(my)]),
                                max([self.ylim[1], np.max(my)])
                            ]

                    if self.screen.isChecked() and not (self.parent.visual_stim
                                                        is not None):
                        print('initializing stim [...]')
                        self.parent.load_VisualStim()
                    if self.screen.isChecked():
                        inset = ge.inset(AX[irow][icol], [.8, .9, .3, .25])
                        self.parent.visual_stim.show_frame(\
                                    self.EPISODES['index_from_start'][cond][0],
                                    ax=inset, parent=self.parent, enhance=True, label=None)

        if self.withStatTest.isChecked():
            for irow, row_cond in enumerate(ROW_CONDS):
                for icol, col_cond in enumerate(COL_CONDS):
                    for icolor, color_cond in enumerate(COLOR_CONDS):

                        cond = np.array(
                            single_cond & col_cond & row_cond
                            & color_cond)[:self.EPISODES['resp'].shape[0]]
                        test = stat_test_for_evoked_responses(
                            self.EPISODES,
                            cond,
                            interval_pre=[self.t0pre, self.t1pre],
                            interval_post=[self.t0post, self.t1post],
                            test='wilcoxon')

                        AX[irow][icol].plot([self.t0pre, self.t1pre],
                                            self.ylim[0] * np.ones(2),
                                            'k-',
                                            lw=2)
                        AX[irow][icol].plot([self.t0post, self.t1post],
                                            self.ylim[0] * np.ones(2),
                                            'k-',
                                            lw=2)
                        ps, size = pval_to_star(test)
                        AX[irow][icol].annotate(
                            ps,
                            ((self.t1pre + self.t0post) / 2., self.ylim[0]),
                            va='top',
                            ha='center',
                            size=size,
                            xycoords='data')

        if (self.ymin.text() != '') and (self.ymax.text() != ''):
            self.ylim = [float(self.ymin.text()), float(self.ymax.text())]
        if (self.xmin.text() != '') and (self.xmax.text() != ''):
            self.xlim = [float(self.xmin.text()), float(self.xmax.text())]
        else:
            self.xlim = [self.EPISODES['t'][0], self.EPISODES['t'][-1]]

        for irow, row_cond in enumerate(ROW_CONDS):
            for icol, col_cond in enumerate(COL_CONDS):
                ge.set_plot(AX[irow][icol],
                            spines=(['left', 'bottom']
                                    if self.axis.isChecked() else []),
                            ylim=self.ylim,
                            xlim=self.xlim,
                            xlabel=(self.xbarlabel.text()
                                    if self.axis.isChecked() else ''),
                            ylabel=(self.ybarlabel.text()
                                    if self.axis.isChecked() else ''))

                if self.stim.isChecked():
                    AX[irow][icol].fill_between(
                        [0, np.mean(self.EPISODES['time_duration'])],
                        self.ylim[0] * np.ones(2),
                        self.ylim[1] * np.ones(2),
                        color='grey',
                        alpha=.2,
                        lw=0)

        if not self.axis.isChecked():
            ge.draw_bar_scales(AX[0][0],
                               Xbar=(0. if self.xbar.text() == '' else float(
                                   self.xbar.text())),
                               Xbar_label=self.xbarlabel.text(),
                               Ybar=(0. if self.ybar.text() == '' else float(
                                   self.ybar.text())),
                               Ybar_label=self.ybarlabel.text(),
                               Xbar_fraction=0.1,
                               Xbar_label_format='%.1f',
                               Ybar_fraction=0.2,
                               Ybar_label_format='%.1f',
                               loc='top-left')

        if self.label.text() != '':
            ge.annotate(fig, ' '+self.label.text()+\
                        (1+int(self.nlabel.text()))*'\n', (0,0), color=COLORS[0],
                        ha='left', va='bottom')

        if self.annot.isChecked():
            S = ''
            if hasattr(self, 'roiPick'):
                S += 'roi #%s' % self.roiPick.text()
            for i, key in enumerate(self.varied_parameters.keys()):
                if 'single-value' in getattr(self,
                                             '%s_plot' % key).currentText():
                    S += ', %s=%.2f' % (key, getattr(
                        self, '%s_values' % key).currentText())
            ge.annotate(fig, S, (0, 0), color='k', ha='left', va='bottom')

        return fig, AX

    def plot_raw_data(self):

        self.fig, ax = ge.figure(axes_extents=[[[6, 4]]],
                                 bottom=0.1,
                                 right=3.,
                                 top=0.5)

        Tbar = 1  # by default
        if (self.xmin.text() == '') or (self.xmax.text() == ''):
            tlim = [0, 20]
        else:
            tlim = [float(self.xmin.text()), float(self.xmax.text())]

        if self.xbar.text() != '':
            Tbar = float(self.xbar.text())

        settings = {}
        for mod in self.modalities:
            if getattr(self, mod + 'Plot').value() > 0:
                settings[mod] = dict(fig_fraction=getattr(self, mod +
                                                          'Plot').value())

        for key in ['CaImaging', 'CaImagingSum']:
            if (key in settings) and (self.sqbox.text() in [
                    'dF/F', 'Fluorescence', 'Neuropil', 'Deconvolved'
            ]):
                settings[key]['subquantity'] = self.sqbox.text()
        if 'CaImaging' in settings:
            settings['CaImaging']['roiIndices'] = self.parent.roiIndices

        self.data.plot(tlim, settings=settings, Tbar=Tbar, ax=ax)
        ge.show()

    def plot_FOV(self):

        if self.fovMap.currentText() == 'black_to_green':
            cmap = ge.get_linear_colormap('k', ge.green)
        elif self.fovMap.currentText() == 'white_to_green':
            cmap = ge.get_linear_colormap('w', 'darkgreen')
        elif self.fovMap.currentText() == 'grey':
            cmap = ge.get_linear_colormap('w', 'darkgrey')
        else:
            cmap = ge.viridis

        self.fig, ax = ge.figure(figsize=(2., 4.),
                                 bottom=0,
                                 top=0,
                                 right=0.,
                                 left=0.)
        self.data.show_CaImaging_FOV(\
                                    self.fovType.currentText(),
                                    NL=float(self.fovNL.value()), cmap=cmap, ax=ax)
        ge.annotate(ax,
                    self.fovLabel.text(), (.99, .99),
                    ha='right',
                    va='top',
                    color='white')
        ge.show()

    def plot_correl(self):
        pass

    def plot(self):

        if self.responseType.currentText() == 'stim-evoked-traces':
            self.fig, AX = self.plot_row_column_of_quantity()
        else:
            pass

        ge.show()

    def select_ROI(self):
        """ see dataviz/gui.py   //  dataviz/guiparts.py """
        roiIndices = self.select_ROI_from_pick(cls=self.parent)
        if len(roiIndices) > 0:
            self.parent.roiIndices = roiIndices
            self.parent.roiPick.setText(self.roiPick.text())
        self.statusBar.showMessage('ROIs set to %s ' % self.parent.roiIndices)

    def compute_episodes(self):

        if self.sqbox.text() in [
                'dF/F', 'Fluorescence', 'Neuropil', 'Deconvolved'
        ]:
            self.parent.CaImaging_key = self.sqbox.text()

        self.t0pre = float(self.preWindow.text().replace('[', '').replace(
            ']', '').split(',')[0])
        self.t1pre = float(self.preWindow.text().replace('[', '').replace(
            ']', '').split(',')[1])
        self.t0post = float(self.postWindow.text().replace('[', '').replace(
            ']', '').split(',')[0])
        self.t1post = float(self.postWindow.text().replace('[', '').replace(
            ']', '').split(',')[1])

        if (self.qbox.currentIndex() > 0) and (self.pbox.currentIndex() > 0):
            self.EPISODES = build_episodes(
                self,
                parent=self.parent,
                protocol_id=self.pbox.currentIndex() - 1,
                quantity=self.qbox.currentText(),
                dt_sampling=self.samplingBox.value(),  # ms
                interpolation='linear',
                baseline_substraction=self.baseline.isChecked(),
                prestim_duration=(self.t1pre - self.t0pre),
                verbose=True)
        else:
            print(' /!\ Pick a protocol and a quantity')

    def build_conditions(self, X, K):
        if len(K) > 0:
            CONDS = []
            XK = np.meshgrid(*X)
            for i in range(len(
                    XK[0].flatten())):  # looping over joint conditions
                cond = np.ones(np.sum(self.Pcond), dtype=bool)
                for k, xk in zip(K, XK):
                    cond = cond & (self.parent.nwbfile.stimulus[k].data[
                        self.Pcond] == xk.flatten()[i])
                CONDS.append(cond)
            return CONDS
        else:
            return [np.ones(np.sum(self.Pcond), dtype=bool)]

    def build_single_conditions(self):

        full_cond = np.ones(np.sum(self.Pcond), dtype=bool)

        for i, key in enumerate(self.varied_parameters.keys()):
            if 'single-value' in getattr(self, '%s_plot' % key).currentText():
                cond = (np.array(
                    self.parent.nwbfile.stimulus[key].data[self.Pcond],
                    dtype=str) != getattr(self,
                                          '%s_values' % key).currentText())
                full_cond[cond] = False

        return full_cond

    def build_column_conditions(self):
        X, K = [], []
        for i, key in enumerate(self.varied_parameters.keys()):
            if 'column' in getattr(self, '%s_plot' % key).currentText():
                X.append(
                    np.sort(
                        np.unique(self.parent.nwbfile.stimulus[key].data[
                            self.Pcond])))
                K.append(key)
        return self.build_conditions(X, K)

    def build_row_conditions(self):
        X, K = [], []
        for i, key in enumerate(self.varied_parameters.keys()):
            if 'row' in getattr(self, '%s_plot' % key).currentText():
                X.append(
                    np.sort(
                        np.unique(self.parent.nwbfile.stimulus[key].data[
                            self.Pcond])))
                K.append(key)
        return self.build_conditions(X, K)

    def build_color_conditions(self):
        X, K = [], []
        for i, key in enumerate(self.varied_parameters.keys()):
            if 'color' in getattr(self, '%s_plot' % key).currentText():
                X.append(
                    np.sort(
                        np.unique(self.parent.nwbfile.stimulus[key].data[
                            self.Pcond])))
                K.append(key)
        return self.build_conditions(X, K)
Пример #3
0
def make_sumary_pdf(filename,
                    Nmax=1000000,
                    include=['exp', 'rois', 'raw', 'protocols'],
                    T_raw_data=60,
                    N_raw_data=5,
                    ROI_raw_data=20,
                    Tbar_raw_data=5):

    data = MultimodalData(filename)
    data.roiIndices = np.sort(
        np.random.choice(np.arange(data.iscell.sum()),
                         size=min([data.iscell.sum(), ROI_raw_data]),
                         replace=False))

    folder = filename.replace('.nwb', '')
    if not os.path.isdir(folder):
        os.mkdir(folder)

    if 'exp' in include:
        with PdfPages(os.path.join(folder, 'exp.pdf')) as pdf:

            print('writing experimental metadata ')
            fig = metadata_fig(data)
            pdf.savefig()  # saves the current figure into a pdf page
            plt.close()

            print('plotting behavior analysis ')
            fig = behavior_analysis_fig(data)
            pdf.savefig()  # saves the current figure into a pdf page
            plt.close()

    if 'exp' in include:
        with PdfPages(os.path.join(folder, 'rois.pdf')) as pdf:

            print('plotting imaging FOV ')
            fig, AX = plt.subplots(1, 4, figsize=(11.4, 2))
            data.show_CaImaging_FOV(key='meanImg',
                                    NL=1,
                                    cmap='viridis',
                                    ax=AX[0])
            data.show_CaImaging_FOV(key='meanImg',
                                    NL=2,
                                    cmap='viridis',
                                    ax=AX[1])
            data.show_CaImaging_FOV(key='meanImgE',
                                    NL=2,
                                    cmap='viridis',
                                    ax=AX[2])
            data.show_CaImaging_FOV(key='max_proj',
                                    NL=2,
                                    cmap='viridis',
                                    ax=AX[3])
            pdf.savefig()  # saves the current figure into a pdf page
            plt.close()

    if 'raw' in include:

        with PdfPages(os.path.join(folder, 'raw.pdf')) as pdf:

            print('plotting full data view [...]')
            fig, ax = plt.subplots(1, figsize=(11.4, 5))
            fig.subplots_adjust(top=0.8, bottom=0.05)
            data.plot(data.tlim,
                      settings=raw_data_plot_settings(data,
                                                      subsampling_factor=1000),
                      ax=ax,
                      Tbar=Tbar_raw_data)
            pdf.savefig()  # saves the current figure into a pdf page
            plt.close()

            # # plot raw data sample
            # for t0 in np.linspace(T_raw_data, data.tlim[1], N_raw_data):
            #     TLIM = [np.max([10,t0-T_raw_data]),t0]
            #     print('plotting raw data sample at times ', TLIM)
            #     fig, ax = plt.subplots(1, figsize=(11.4, 5))
            #     fig.subplots_adjust(top=0.8, bottom=0.05)
            #     data.plot(TLIM, settings=raw_data_plot_settings(data),
            #               ax=ax, Tbar=Tbar_raw_data)
            #     # inset with time sample
            #     axT = plt.axes([0.6, 0.9, 0.3, 0.05])
            #     axT.axis('off')
            #     axT.plot(data.tlim, [0,0], 'k-', lw=2)
            #     axT.plot(TLIM, [0,0], '-', color=plt.cm.tab10(3), lw=5)
            #     axT.annotate('0 ', (0,0), xycoords='data', ha='right', fontsize=9)
            #     axT.annotate(' %.1fmin' % (data.tlim[1]/60.), (data.tlim[1],0), xycoords='data', fontsize=9)
            #     pdf.savefig()  # saves the current figure into a pdf page
            #     plt.close()

    print('looping over protocols for analysis [...]')

    if 'protocols' in include:
        # looping over protocols
        for p, protocol in enumerate(data.protocols):

            print('plotting protocol "%s" [...]' % protocol)

            with PdfPages(os.path.join(folder, '%s.pdf' % protocol)) as pdf:

                # finding protocol type
                protocol_type = (data.metadata['Protocol-%i-Stimulus' %
                                               (p + 1)] if
                                 (len(data.protocols) > 1) else
                                 data.metadata['Stimulus'])
                print(protocol_type)
                # then protocol-dependent analysis

                if protocol_type == 'full-field-grating':
                    from analysis.orientation_direction_selectivity import orientation_selectivity_analysis
                    Nresp, SIs = 0, []
                    for i in range(data.iscell.sum())[:Nmax]:
                        fig, SI, responsive = orientation_selectivity_analysis(
                            data, roiIndex=i, verbose=False)
                        pdf.savefig(
                        )  # saves the current figure into a pdf page
                        plt.close()
                        if responsive:
                            Nresp += 1
                            SIs.append(SI)
                    # summary figure for this protocol
                    fig, AX = summary_fig(Nresp, data.iscell.sum(),
                                          np.array(SIs))
                    pdf.savefig()  # saves the current figure into a pdf page
                    plt.close()

                elif protocol_type == 'drifting-full-field-grating':
                    from analysis.orientation_direction_selectivity import direction_selectivity_analysis
                    Nresp, SIs = 0, []
                    for i in range(data.iscell.sum())[:Nmax]:
                        fig, SI, responsive = direction_selectivity_analysis(
                            data, roiIndex=i, verbose=False)
                        pdf.savefig(
                        )  # saves the current figure into a pdf page
                        plt.close()
                        if responsive:
                            Nresp += 1
                            SIs.append(SI)
                    fig, AX = summary_fig(Nresp,
                                          data.iscell.sum(),
                                          np.array(SIs),
                                          label='Direction Select. Index')
                    pdf.savefig()  # saves the current figure into a pdf page
                    plt.close()

                elif protocol_type in [
                        'center-grating', 'drifting-center-grating'
                ]:
                    from surround_suppression import orientation_size_selectivity_analysis
                    Nresp, SIs = 0, []
                    for i in range(data.iscell.sum())[:Nmax]:
                        fig, responsive = orientation_size_selectivity_analysis(
                            data, roiIndex=i, verbose=False)
                        pdf.savefig(
                        )  # saves the current figure into a pdf page
                        plt.close()
                        if responsive:
                            Nresp += 1
                            SIs.append(0)  # TO BE FILLED
                    fig, AX = summary_fig(Nresp,
                                          data.iscell.sum(),
                                          np.array(SIs),
                                          label='none')
                    pdf.savefig()  # saves the current figure into a pdf page
                    plt.close()

                elif 'noise' in protocol_type:
                    from receptive_field_mapping import RF_analysis
                    Nresp, SIs = 0, []
                    for i in range(data.iscell.sum())[:Nmax]:
                        fig, SI, responsive = RF_analysis(data,
                                                          roiIndex=i,
                                                          verbose=False)
                        pdf.savefig(
                        )  # saves the current figure into a pdf page
                        plt.close()
                        if responsive:
                            Nresp += 1
                            SIs.append(0)  # TO BE FILLED
                    fig, AX = summary_fig(Nresp,
                                          data.iscell.sum(),
                                          np.array(SIs),
                                          label='none')
                    pdf.savefig()  # saves the current figure into a pdf page
                    plt.close()

                elif 'spatial-location' in protocol_type:
                    from surround_suppression import orientation_size_selectivity_analysis
                    Nresp, SIs = 0, []
                    for i in range(data.iscell.sum())[:Nmax]:
                        fig, responsive = orientation_size_selectivity_analysis(
                            data, roiIndex=i, verbose=False)
                        pdf.savefig(
                        )  # saves the current figure into a pdf page
                        plt.close()
                        if responsive:
                            Nresp += 1
                            SIs.append(0)  # TO BE FILLED
                    fig, AX = summary_fig(Nresp,
                                          data.iscell.sum(),
                                          np.array(SIs),
                                          label='none')
                    pdf.savefig()  # saves the current figure into a pdf page
                    plt.close()

    print('[ok] pdfs succesfully saved in "%s" !' % folder)