def __init__(self, parent):
wx.Dialog.__init__(self, parent, wx.ID_ANY, 'Cluster Center Chooser', size=(550, 550))
self.CenterOnParent()
self.centers = []
# set up the plotting panel
self.dataPanel = CenterChooserPanel(self, self.OnClick)
(self.dataSelectors, self.selectorSizer) = customsizers.ChannelSelectorSizer(self)
if DataStore.getCurrentIndex() is not None:
self.dataPanel.addData(DataStore.getCurrentIndex())
# filter the list of labels to include only user-selected ones
selDims = DataStore.getCurrentDataSet().selDims
labels = [DataStore.getCurrentDataSet().labels[i] for i in selDims]
customsizers.populateSelectors(self.dataSelectors, labels, selDims)
self.dataPanel.updateAxes((selDims[0],selDims[1]))
# main sizer
self.sizer = wx.BoxSizer(wx.VERTICAL)
self.sizer.Add(self.dataPanel, True, wx.EXPAND | wx.BOTTOM | wx.TOP, -20)
self.sizer.AddSpacer(20)
self.sizer.Add(self.selectorSizer, False, wx.EXPAND | wx.TOP, 20)
self.sizer.Add(self.CreateButtonSizer(wx.OK | wx.CANCEL), False, wx.EXPAND)
self.SetSizer(self.sizer)
self.SetBackgroundColour("white")
def isolateClusters(selection, datasetName):
"""
Create a new data set from a selection of clusters.
@type selection: list
@var selection: The indices of the clusters to be merged.
@type datasetName: str
@var datasetName: The display name for the new data set.
"""
if (len(selection) > 0):
currFCData = DataStore.getCurrentDataSet()
if (datasetName == ""):
datasetName = currFCData.displayname
clusters = separate(currFCData.data, currFCData.getCurrentClustering())
# merge selected clusters
newData = dh.mergeData(selection, clusters)
# assign new data set to the store
newFCData = FacsData('', currFCData.labels, newData, parent=currFCData.ID)
newFCData.displayname = datasetName
newFCData.selDims = currFCData.selDims
# add basic text annotations
textAnn = {'parent': currFCData.displayname}
textAnn['events'] = len(newFCData.data)
newFCData.annotations['text'] = textAnn
DataStore.add(newFCData)
def AddChosenCenter(self, point):
cbxSelected = self._GetSelectedAxes()
dims = DataStore.getCurrentDataSet().labels
if (DataStore.getCurrentDataSet().selDims):
dims = DataStore.getCurrentDataSet().selDims
retPoint = [0 for i in dims]
for i in range(len(cbxSelected)):
retPoint[cbxSelected[i]] = point[i]
self.centers.append(retPoint)
self.dataPanel.addCenter(retPoint)
def addSubplot(self):
cds = DataStore.getCurrentDataSet()
if cds is None:
return
try:
clusteringIndex = cds.clustering.keys()[-1]
except IndexError:
self.facsPlotPanel.addSubplot(DataStore.getCurrentIndex())
else:
self.facsPlotPanel.addSubplot(DataStore.getCurrentIndex(), clusteringIndex)
def OnExport(self, event):
if DataStore.getCurrentIndex() is None:
wx.MessageBox("Please load a dataset before attempting to export",
"Data Missing", wx.OK | wx.ICON_ERROR)
return
format = io.getMethod(event.GetId())().FileType
dlg = wx.FileDialog(self, "Save File", "", "",
format,
wx.FD_SAVE|wx.FD_OVERWRITE_PROMPT|wx.FD_CHANGE_DIR)
if dlg.ShowModal() == wx.ID_OK:
io.exportDataFile(event.GetId(), dlg.Path, DataStore.getCurrentDataSet(), window=self)
dlg.Destroy()
def deleteSelection(self):
"""
Delete the data object referenced by the current tree selection.
"""
item = self.getSanitizedItemSelectionData()
if item is None:
return
if item[0] is DATA_SET_ITEM:
ids = (item[1], None) if len(item) < 3 else (item[1], item[2])
self.Parent.TopLevelParent.facsPlotPanel.deleteAssociatedSubplots(ids)
self.applyToSelection(DataStore.remove, FacsData.removeClustering)
# if all data deleted, clear axes selectors
if len(DataStore.getData()) == 0:
self.Parent.TopLevelParent.updateAxesList([])
if item[0] is FIGURE_SET_ITEM:
id = item[1]
#TODO: figure out if this is a good shortcut for clearing the figure of subplots
if id == FigureStore.getSelectedIndex():
wx.MessageBox('The currently selected Figure cannot be deleted.', 'Invalid Action', wx.OK | wx.ICON_WARNING)
else:
FigureStore.remove(id)
self.updateTree()
def updateTree(self):
"""
Rebuilds the tree from the current state of the DataStore.
"""
self.tree.DeleteAllItems()
self.root = self.tree.AddRoot("Project")
self.tree.SetPyData(self.root, None)
self.tree.SetItemImage(self.root, self.fldridx, wx.TreeItemIcon_Normal)
self.tree.SetItemImage(self.root, self.fldropenidx, wx.TreeItemIcon_Expanded)
dataTree = self.tree.AppendItem(self.root, "Data Sets")
self.tree.SetPyData(dataTree, DATA_SET_ITEM)
self.tree.SetItemImage(dataTree, self.fldridx, wx.TreeItemIcon_Normal)
self.tree.SetItemImage(dataTree, self.fldropenidx, wx.TreeItemIcon_Expanded)
figsTree = self.tree.AppendItem(self.root, "Figure Sets")
self.tree.SetPyData(figsTree, FIGURE_SET_ITEM)
self.tree.SetItemImage(figsTree, self.fldridx, wx.TreeItemIcon_Normal)
self.tree.SetItemImage(figsTree, self.fldropenidx, wx.TreeItemIcon_Expanded)
# create tree
self.buildDataTree(dataTree, DataStore.getData(), [])
self.buildFigureTree(figsTree, FigureStore.getFigures())
self.tree.Expand(self.root)
if self.dataTreeExpanded:
self.tree.Expand(dataTree)
if self.figureTreeExpanded:
self.tree.Expand(figsTree)
def __init__(self, parent):
wx.Dialog.__init__(self, parent, wx.ID_ANY, "Select clusterings to match colors",
style=wx.RESIZE_BORDER|wx.DEFAULT_DIALOG_STYLE, size=(400, 150))
self.CenterOnParent()
allData = DataStore.getData()
self.choices = []
self.choiceIDs = []
# Populate the choices list with string, and populate the
# choiceIDs list with (dataID, clusteringID) tuples so the
# combo box selection can be tied to the data
for didx in allData:
fcData = allData[didx]
for cidx in fcData.clustering:
self.choices.append(fcData.displayname + ": " +
getStringRepr(fcData.methodIDs[cidx]) + " " +
str(cidx+1))
self.choiceIDs.append((fcData.ID, cidx))
self.cbxSourceClustering = wx.ComboBox(self, choices=self.choices, style=wx.CB_READONLY)
self.cbxDestClustering = wx.ComboBox(self, choices=self.choices, style=wx.CB_READONLY)
self.formSizer = wx.FlexGridSizer(2, 2, vgap=5, hgap=5)
self.formSizer.FlexibleDirection = wx.HORIZONTAL
self.formSizer.AddF(wx.StaticText(self, -1, 'First Clustering:'), wx.SizerFlags(1).Expand())
self.formSizer.AddF(self.cbxSourceClustering, wx.SizerFlags(2).Expand())
self.formSizer.AddF(wx.StaticText(self, -1, 'Second Clustering:'), wx.SizerFlags(1).Expand())
self.formSizer.AddF(self.cbxDestClustering, wx.SizerFlags(2).Expand())
self.Sizer = wx.BoxSizer(wx.VERTICAL)
self.Sizer.AddF(self.formSizer, wx.SizerFlags(1).Expand().Border(wx.ALL, 10))
self.Sizer.AddF(self.CreateButtonSizer(wx.OK|wx.CANCEL), wx.SizerFlags().Expand().Border(wx.BOTTOM, 10))
def renameItem(self):
"""
Give a new display name to the currently selected data item
"""
item = self.getSanitizedItemSelectionData()
if item is None:
return
if (item[0] is DATA_SET_ITEM):
data = DataStore.get(item[1])
dlg = EditNameDialog(self.Parent, data.displayname)
if dlg.ShowModal() == wx.ID_OK:
data.displayname = dlg.Text
dlg.Destroy()
if (item[0] is FIGURE_SET_ITEM):
figure = FigureStore.get(item[1])
dlg = EditNameDialog(self.Parent, figure.name)
if dlg.ShowModal() == wx.ID_OK:
figure.name = dlg.Text
dlg.Destroy()
item = self.tree.GetSelection()
item.SetText(dlg.Text)
self.tree.RefreshSelected()
item.SetHilight(False)
item.SetHilight(True)
def setExpanded(self, item, flag):
#item = self.getItemSelectionData()
data = self.tree.GetItemPyData(item)
if isinstance(data, int):
if data is DATA_SET_ITEM:
self.dataTreeExpanded = flag
if data is FIGURE_SET_ITEM:
self.figureTreeExpanded = flag
if isinstance(data, tuple):
if data[0] is DATA_SET_ITEM:
if len(data) == 2:
DataStore.getData()[data[1]].nodeExpanded = flag
if len(data) > 3:
DataStore.getData()[data[1]].infoExpanded[data[2]] = flag
def OnCluster(self, event):
"""
Handles all clustering requests.
Clustering requests are handled in the following method:
1. Passes the requested clustering method to
L{cluster.dialogs.ClusterOptionsDialog.showDialog} method
2. Passes the data and the returned method options to L{cluster.methods.cluster}
3. Passes the returned cluster membership list to the FacsPlotPanel for display
"""
dlg = cDlgs.getClusterDialog(event.GetId(), self)
if dlg.ShowModal() == wx.ID_OK:
if (DataStore.getCurrentDataSet() is not None):
self.statusbar.SetStatusText('Running %s clustering...' % cMthds.methods[event.GetId()][1], 0)
fcs = DataStore.getCurrentDataSet()
data = fcs.data
# Remove columns from analysis as specified by the user
if len(fcs.selDims) > 0:
data = dh.filterData(data, fcs.selDims)
clusterIDs, msg = cMthds.cluster(event.GetId(), data, **dlg.getMethodArgs())
DataStore.addClustering(event.GetId(), clusterIDs, dlg.getMethodArgs())
clusteringIndex = DataStore.getCurrentDataSet().clustering.keys()[-1]
self.statusbar.SetStatusText(msg, 0)
if (dlg.isApplyChecked()):
if self.facsPlotPanel.SelectedSubplotIndex is not None:
self.facsPlotPanel.CurrentSubplot = dv.Subplot(self.facsPlotPanel.SelectedSubplotIndex,
DataStore.getCurrentIndex(), clusteringIndex)
self.facsPlotPanel.draw()
else:
self.facsPlotPanel.addSubplot(DataStore.getCurrentIndex(), clusteringIndex)
self.treeCtrlPanel.updateTree()
dlg.Destroy()
class Analyzer:
def __init__(self, games, previous_length):
self.previous_length = previous_length
self.data_store = DataStore(previous_length)
self.games = [Attributes(game) for game in games]
def run(self):
games_dict = {}
for game in self.games:
away_team = self.data_store.get_data(game.away_team)
home_team = self.data_store.get_data(game.home_team)
if not away_team or not home_team:
self.add_game_data(game)
continue
enough_away_stats = len(
away_team.previous_stats) == self.previous_length
enough_home_stats = len(
home_team.previous_stats) == self.previous_length
enough_previous_stats = enough_away_stats and enough_home_stats
if not enough_previous_stats:
self.add_game_data(game)
continue
away_pred = predict_score(away_team, home_team)
home_pred = predict_score(home_team, away_team)
games_dict[game.id] = (away_pred, home_pred)
return games_dict
def add_game_data(self, game):
self.data_store.add_data(game.away_team)
self.data_store.add_data(game.home_team)
for player in game.away_players:
self.data_store.add_data(player)
for player in game.home_players:
self.data_store.add_data(player)
def OnLoadState(self, event):
from data.io import loadState
if len(DataStore.getData()) > 0:
dlgWarn = wx.MessageDialog(self, 'This action may overwrite currently loaded datasets and/or plots.\n\nContinue anyway?', 'Warning',
wx.YES_NO | wx.NO_DEFAULT | wx.ICON_WARNING)
if dlgWarn.ShowModal() == wx.ID_NO:
dlgWarn.Destroy()
return
dlgWarn.Destroy()
formats = "FIND Project File (*.find)|*.find"
dlg = wx.FileDialog(self, "Select saved project", self.dirname, "", formats, wx.FD_OPEN)
if dlg.ShowModal() == wx.ID_OK:
try:
loadState(dlg.Directory, dlg.Filename)
except error.ProjectLoadingError:
return
# Load all Figures with Subplot instances from the stored dicts
for fID in FigureStore.getFigures():
fig = FigureStore.get(fID)
splots = []
for plot in fig.subplots:
s = dv.Subplot()
s.load(plot)
s.parent = self.facsPlotPanel.figure
splots.append(s)
fig.subplots = splots
currFigure = FigureStore.getSelectedFigure()
self.facsPlotPanel.subplots = currFigure.subplots
self.facsPlotPanel.SelectedSubplotIndex = currFigure.selectedSubplot
self.facsPlotPanel.updateAxes(currFigure.axes, False)
self.facsPlotPanel.updateSubplotGrid(currFigure.grid[0], currFigure.grid[1], True)
self.chkLinked.Value = self.facsPlotPanel.CurrentSubplotLinked
labels = DataStore.getCurrentDataSet().labels if DataStore.getCurrentDataSet() is not None else []
self.updateAxesList(labels, currFigure.axes)
self.treeCtrlPanel.updateTree()
self.statusbar.SetStatusText("Project loaded from %s" % dlg.Path, 0)
dlg.Destroy()
def OnEditChannelNames(self, event):
"""
Allow users to modify the descriptors for each channel (dimension)
"""
fcData = DataStore.getCurrentDataSet()
if fcData is not None:
dgridDlg = displayDialogs.SampleDataDisplayDialog(self, fcData.data[0:10,:],
fcData.labels, 'Edit Channel Labels',
False, False)
if (dgridDlg.ShowModal() == wx.ID_OK):
# reassign FacsData instance labels
for fcd in DataStore.getData().values():
labels = [item[1] for item in sorted(dgridDlg.ColumnLabels.iteritems(), key=itemgetter(0))]
fcd.labels = labels
# update channel selection cbxs
self.updateAxesList(labels, self.facsPlotPanel.SelectedAxes)
# refresh plot window
self.facsPlotPanel.draw()
dgridDlg.Destroy()
else:
wx.MessageBox("There are no data currently loaded.", "Error", wx.OK | wx.ICON_ERROR)
def reassignClusterIDs(src, dst):
"""
Given the cluster centers for two clusterings, determine the centers most
similar to each other and reassign the cluster ids to match.
"""
srcFCS = DataStore.getData()[src[0]]
dstFCS = DataStore.getData()[dst[0]]
srcdata = srcFCS.data
if srcFCS.selDims:
srcdata = dh.filterData(srcFCS.data, srcFCS.selDims)
srcids = srcFCS.clustering[src[1]]
srccenters = pc.clustercentroids(srcdata, clusterid=srcids)[0]
dstdata = dstFCS.data
if dstFCS.selDims:
dstdata = dh.filterData(dstFCS.data, dstFCS.selDims)
dstids = dstFCS.clustering[dst[1]]
dstcenters = pc.clustercentroids(dstdata, clusterid=dstids)[0]
srcsep = separate(srcdata, srcids)
dstsep = separate(dstdata, dstids)
centerEQ = {}
taken = []
# Fill the map with the closest source center for each destination center
for i,dc in enumerate(dstcenters):
bestDist = -1
for j,sc in enumerate(srccenters):
if (j not in taken):
dist = nonSymmetricClusterDistance(dstsep[i], srcsep[j])
if (bestDist < 0) or (dist < bestDist):
bestDist = dist
centerEQ[i] = j
taken.append(centerEQ[i])
# Renumber the cluster IDs in the destination to match the IDs of the closest src center
tmp = [centerEQ[id] for id in dstids]
DataStore.getData()[dst[0]].clustering[dst[1]] = tmp
def scatterplot2D(subplot, figure, dims):
# set default plot options if necessary
opts = subplot.opts
if len(opts) == 0:
opts['xRange'] = ()
opts['xRangeAuto'] = True
opts['yRange'] = ()
opts['yRangeAuto'] = True
opts['xTransform'] = ''
opts['yTransform'] = ''
opts['transformAuto'] = True
# Set axes transforms
if (opts['transformAuto']):
fcData = DataStore.get(subplot.dataIndex)
tf = ('xTransform', 'yTransform')
for i, dim in enumerate(dims):
t = fcData.getDefaultTransform(dim)
if (t is not None) and ('log' in t.lower()):
opts[tf[i]] = 'log'
else:
opts[tf[i]] = 'linear'
if opts['xRangeAuto']:
opts['xRange'] = (1, np.max(subplot.Data[:,dims[0]])*1.5)
if opts['yRangeAuto']:
opts['yRange'] = (1, np.max(subplot.Data[:,dims[1]])*1.5)
# create the subplot and set its attributes
subplot.axes = figure.add_subplot(subplot.mnp, xlim=opts['xRange'],
ylim=opts['yRange'], autoscale_on=False)
subplot.axes.set_xscale(opts['xTransform'], nonposx='clip')
subplot.axes.set_yscale(opts['yTransform'], nonposy='clip')
subplot.axes.set_xlabel(subplot.Labels[dims[0]])
subplot.axes.set_ylabel(subplot.Labels[dims[1]])
subplot.axes.set_title(subplot.Title)
# draw the supplied FACS data
if (not subplot.isDataClustered()):
subplot.axes.plot(subplot.Data[:,dims[0]],
subplot.Data[:,dims[1]],
'.', ms=1, color='black')
else:
data = separate(subplot.Data, subplot.Clustering)
for i in range(len(data)):
xs = data[i][:,dims[0]]
ys = data[i][:,dims[1]]
subplot.axes.plot(xs, ys, '.', ms=1, color=methods.plotColors[i])
def buildDataTree(self, parent, dataDict, visited):
for dIndex, fData in dataDict.iteritems():
# add tree node for the FACS data set
if fData.ID in visited:
return
child = self.tree.AppendItem(parent, fData.displayname)
if (DataStore.getCurrentIndex() == dIndex):
self.tree.SetItemBold(child, True)
self.tree.SetPyData(child, (DATA_SET_ITEM, dIndex))
self.tree.SetItemImage(child, self.fldridx, wx.TreeItemIcon_Normal)
self.tree.SetItemImage(child, self.fldropenidx, wx.TreeItemIcon_Expanded)
# add nodes for all the clusterings of the current data set
for cIndex in fData.clustering:
clust = self.tree.AppendItem(child,
cluster.methods.getStringRepr(fData.methodIDs[cIndex]) +
" " + str(cIndex+1))
if ((fData.selectedClustering == cIndex) and (fData.ID == DataStore.getCurrentIndex())):
self.tree.SetItemBold(clust, True)
toolTip = clusteringInfo(fData, cIndex)
self.tree.SetPyData(clust, (DATA_SET_ITEM, dIndex, cIndex, toolTip)) #(data index, cluster index, tooltip)
self.tree.SetItemImage(clust, self.fldridx, wx.TreeItemIcon_Normal)
self.tree.SetItemImage(clust, self.fldropenidx, wx.TreeItemIcon_Expanded)
# set clustering options
for line in toolTip.split('\n'):
if (line != ''):
info = self.tree.AppendItem(clust, line)
self.tree.SetPyData(info, None)
if (fData.infoExpanded[cIndex]):
self.tree.Expand(clust)
# recursive call to build the tree for any children
visited.append(fData.ID)
self.buildDataTree(child, dict([(i,DataStore.getData()[i]) for i in fData.children]), visited)
if (fData.nodeExpanded and self.tree.GetChildrenCount(child) > 0):
self.tree.Expand(child)
def OnAnalyze(self, event):
"""
Handles all requests for analysis methods; built-in and plugins.
Currently, analysis methods expect data, a list of dimensions
available to use in analysis, and a window ref in order to
display a subwindow/dialog with results or options.
Analysis methods are expected to return data and/or a status
message. Currently returned data is only used when called from
code, not from the menu; which this method represents.
"""
if (DataStore.getCurrentDataSet() is not None):
strID = aMthds.strID(event.GetId())
self.statusbar.SetStatusText('Running %s...' % aMthds.AvailableMethods()[strID][2], 0)
fcs = DataStore.getCurrentDataSet()
data = fcs.data
# Remove columns from analysis as specified by the user
if len(fcs.selDims) > 0:
data = dh.filterData(data, fcs.selDims)
args = {'parentWindow': self}
_, msg = aMthds.getMethod(strID)(data, **args)
self.statusbar.SetStatusText(msg, 0)
def OnAddFigure(self, event):
if len(DataStore.getData()) == 0:
return
nameDlg = displayDialogs.EditNameDialog(self, '')
if (nameDlg.ShowModal() == wx.ID_OK):
newFig = Figure(nameDlg.Text, [dv.Subplot()], 1, (1,1), (0,1))
currFig = FigureStore.getSelectedFigure()
FigureStore.add(newFig)
dv.switchFigures(self.facsPlotPanel, currFig, newFig)
self.facsPlotPanel.subplots = []
self.addSubplot()
self.treeCtrlPanel.updateTree()
nameDlg.Destroy()
def barplot(subplot, figure, dims=None):
"""
Create a bar chart with one bar for each cluster, and the bar height
equal to the percent of events contained within the cluster.
Some techniques borrowed from:
http://www.scipy.org/Cookbook/Matplotlib/BarCharts
"""
clusters = separate(subplot.Data, subplot.Clustering)
# set default plot options
opts = subplot.opts
if len(opts) == 0:
opts['labelAngle'] = 0 if len(clusters) < 5 else -20
opts['view'] = 'percent'
# correcting for previous version
if 'view' not in opts:
opts['view'] = 'percent'
dataSize = len(subplot.Data)
if opts['view'] == 'toplevel':
dataSize = len(DataStore.getToplevelParent(subplot.dataIndex).data)
if opts['view'] in ['percent', 'toplevel']:
displayNums = [float(len(cluster))/dataSize*100 for cluster in clusters]
else:
displayNums = [len(cluster) for cluster in clusters]
numBars = len(displayNums)
width = 0.5
xlocs = na.array(range(len(displayNums)))+0.5
subplot.axes = figure.add_subplot(subplot.mnp, title=subplot.Title, autoscale_on=False)
subplot.axes.set_xticks(xlocs + width/2)
if opts['view'] == 'counts':
subplot.axes.set_xticklabels(map(lambda x: "%i" % x, displayNums), rotation=opts['labelAngle'])
subplot.axes.set_ylim(0, max(displayNums)*1.2)
else:
subplot.axes.set_xticklabels(map(lambda x: "%.2f%%" % x, displayNums), rotation=opts['labelAngle'])
subplot.axes.set_ylim(0, 100)
subplot.axes.set_xlim(0, xlocs[-1] + width*2)
subplot.axes.xaxis.tick_bottom()
subplot.axes.yaxis.tick_left()
subplot.axes.bar(xlocs, displayNums, width=width, color=methods.plotColors[:numBars])
def applyToSelection(self, dataFunc, clustFunc):
"""
Apply a function to the data selection depending on whether it refers
to a FacsData object or a clustering of a FacsData object.
@type dataFunc: function
@param dataFunc: A function to apply to the FacsData object selected.
@type clustFunc: function
@param clustFunc: A function to apply to the clustering of the FacsData
object selected.
"""
data = self.getSanitizedItemSelectionData()
if data is not None and data[0] is DATA_SET_ITEM:
# determine if item is a data set or a clustering and select it
if len(data) == 2:
dataFunc(data[1])
if len(data) == 3:
clustFunc(DataStore.selectDataSet(data[1]), data[2])
def OnOpen(self, event):
"""
Opens a FACS data file, parses it, and updates the FacsPlotPanel instance.
"""
# retrieve the I/O methods for inputting files
inputMethods = [m[2]() for m in io.AvailableMethods().values()]
formats = '|'.join([m.FileType for m in inputMethods if io.FILE_INPUT in m.register()])
allLabels = []
allColArr = []
allDims = []
fColsMoved = False
numLoaded = 0
# keep track of the common number of dimensions for datasets
numDims = DataStore.getCurrentDataSet()
if numDims is not None:
numDims = len(numDims.labels)
dlg = wx.FileDialog(self, "Choose a file", self.dirname, "", formats,
wx.FD_OPEN|wx.FD_MULTIPLE|wx.FD_CHANGE_DIR)
if dlg.ShowModal() == wx.ID_OK:
#TODO: move to data.io module?
# process each file selected
for n, path in enumerate(dlg.Paths):
self.statusbar.SetStatusText('loading: ' + path, 0)
try:
(labels, data, annotations) = io.loadDataFile(path, window=self)
except TypeError:
# if there was an error loading the file,
# loadDataFile should return None, so skip this file
continue
# make sure the new file matches dimensions of loaded files
if numDims is None:
numDims = len(labels)
elif len(labels) != numDims:
wx.MessageBox("Error loading file: %s\n\nThe number of channels does not match those in currently loaded datasets. \n\nThis file will not be loaded." % dlg.Filenames[n],
"File Error", wx.OK | wx.ICON_ERROR)
continue
# Give the user a brief preview of the data (10 rows) and allow
# column rearrangement and renaming
if (not allLabels):
dgridDlg = displayDialogs.SampleDataDisplayDialog(self, data[0:10,:], labels)
if (dgridDlg.ShowModal() == wx.ID_OK):
ca = dgridDlg.ColumnArrangement
lbls = dgridDlg.ColumnLabels
# Reassign the column labels
labels = [lbls[i] for i in ca]
if dgridDlg.ApplyToAll:
allLabels = list(labels)
allColArr = list(ca)
# Rearrange the data columns
if (dgridDlg.ColumnsMoved):
fColsMoved = True
data = dh.reorderColumns(data, ca)
else:
dgridDlg.Destroy()
continue
dgridDlg.Destroy()
else:
labels = list(allLabels)
if fColsMoved:
data = dh.reorderColumns(data, allColArr)
# update the DataStore
DataStore.add(FacsData(dlg.Filenames[n], labels, data, annotations=annotations))
numLoaded += 1
if n == 0:
self.updateAxesList(labels)
if (not allDims):
# Allow the user to choose columns for use in analysis
dimDlg = displayDialogs.DimensionExclusionDialog(self, labels)
dimDlg.Size=(dimDlg.Size[0]*.75, dimDlg.Size[1]*.8)
if (dimDlg.ShowModal() == wx.ID_OK):
DataStore.getCurrentDataSet().selDims = dimDlg.SelectedDimensions
if (dimDlg.ApplyToAll):
allDims = list(dimDlg.SelectedDimensions)
dimDlg.Destroy()
else:
DataStore.getCurrentDataSet().selDims = list(allDims)
# update the panel
self.facsPlotPanel.updateAxes([0,1])
if (len(self.facsPlotPanel.subplots) == 0 or len(dlg.Paths) > 1):
self.facsPlotPanel.addSubplot(DataStore.getCurrentIndex())
# Create an n/2 x 2 grid for the n selected data files
if (numLoaded > 1):
self.facsPlotPanel.updateSubplotGrid(int(math.ceil(numLoaded/2.0)), 2)
self.statusbar.SetStatusText('All data files loaded.')
if FigureStore.isEmpty():
fig = Figure('Default', self.facsPlotPanel.subplots, 1,
self.facsPlotPanel.Grid,
self.facsPlotPanel.SelectedAxes)
FigureStore.add(fig)
self.treeCtrlPanel.updateTree()
dlg.Destroy()
def displayDataInfo(self):
item = self.getSanitizedItemSelectionData()
if item is not None and item[0] is DATA_SET_ITEM:
data = DataStore.get(item[1])
dlg = DataInfoDialog(self.Parent, data)
dlg.Show()
def clearClusteringSelection(self):
item = self.getSanitizedItemSelectionData()
if item is not None and item[0] is DATA_SET_ITEM:
DataStore.get(item[1]).selectedClustering = None
self.updateTree()
def loadState(dir, filename):
"""
Restore the system state as stored to disk.
@type dir: string
@param path: The directory under which the the saved system state is stored
@type filename: str
@param filename: The name of the saved project file (.find)
@rtype: tuple
@return: A list of subplot settings (dicts) retrieved from the file,
The index of the currently selected subplot,
The currently selected axes,
The number of rows and columns in the figure (grid size)
"""
store = shelve.open(os.path.join(dir, filename))
#store = dbopen(os.path.join(dir, filename))
datakeys = store['data']
try:
bindata = np.load(os.path.join(dir,store['binfile']))
except IOError:
bindata = None
except BadZipfile:
wx.MessageBox('The file \'%s\' may have become corrupted. Project loading has been cancelled' % store['binfile'],
'Data Loading Error',
wx.OK|wx.ICON_ERROR)
raise ProjectLoadingError('BadZipfile: %s' % os.path.join(dir,store['binfile']))
# Parse data sets
for dID in datakeys:
dStr = 'data-%s' % dID
dsett = store[dStr]
ann = dsett['annotations'] if 'annotations' in dsett else {}
ana = dsett['analysis'] if 'analysis' in dsett else {}
fdata = FacsData(dsett['filename'], dsett['labels'], bindata[dStr],
annotations=ann, analysis=ana,
parent=dsett['parent'])
fdata.displayname = dsett['displayname']
fdata.ID = dsett['ID']
fdata.children = dsett['children']
fdata.selDims = dsett['selDims']
fdata.nodeExpanded = dsett['nodeExpanded']
# Parse clusterings
for cID in dsett['clustering']:
cStr = 'clust-%i-%i' % (dID, cID)
csett = store[cStr]
clusterIDs = bindata[cStr]
fdata.addClustering(csett['method'], clusterIDs, csett['opts'], cID)
fdata.clusteringSelDims[cID] = csett['clusteringSelDims']
fdata.infoExpanded[cID] = csett['infoExpanded']
DataStore.add(fdata)
DataStore.selectDataSet(store['current-data'])
# Figures
if 'figures' in store:
for fStr in store['figures']:
fDict = store[fStr]
splots = []
for pStr in fDict['subplots']:
splots.append(store[pStr])
fDict['subplots'] = splots
f = Figure()
f.load(fDict)
FigureStore.add(f)
# handle older save files w/o Figure support
else:
# load the saved subplots into a new 'Default' Figure
plots = []
for pStr in store['plots']:
plots.append(store[pStr])
defFig = Figure('Default', plots, store['current-subplot'], store['grid'], store['selected-axes'])
FigureStore.add(defFig)
if 'current-figure' in store:
FigureStore.setSelectedFigure(store['current-figure'])
else:
FigureStore.setSelectedFigure(0)
def saveState(dir, filename):
"""
Save a representation of the system state: All the loaded data sets,
their clusterings, any transformations or analyses (future),
and all plots.
The state is saved in JSON format based on a dict of the following form:
data: list of IDs
binfile: the filename of the binary file used to store all the actual data
data-dID: dict of settings belonging to a FacsData instance
clust-dID-cID: a dict of attributes belonging to a clustering
figures: list of figureID strings
fig-ID: A dict for each figure keyed on the ID. The subplot attribute here
is replaced with a list of fig-ID-p-ID strings for locating subplot dicts
fig-ID-p-ID: A dict for each subplot in each figure keyed on fig ID and plot ID.
current-data: data ID
current-figure: figure ID
"""
store = shelve.open(os.path.join(dir, filename))
#store = dbopen(os.path.join(dir, filename), 'c', format='csv')
# The actual numeric data will be stored in a separate binary file using
# the numpy savez() method allowing for efficient storage/retrieval of data
binfile = '%s.npz' % filename
bindata = {}
store['data'] = DataStore.getData().keys()
store['binfile'] = binfile
for dID in DataStore.getData():
fdata = DataStore.get(dID)
dStr = 'data-%i' % dID
dfname = fdata.filename if (fdata.filename is not '') else binfile
bindata[dStr] = fdata.data
store[dStr] = {'filename': dfname,
'displayname': fdata.displayname,
'labels': fdata.labels,
'annotations': fdata.annotations,
'analysis': fdata.analysis,
'ID': fdata.ID,
'parent': fdata.parent,
'children': fdata.children,
'selDims': fdata.selDims,
'clustering': fdata.clustering.keys(),
'nodeExpanded': fdata.nodeExpanded,
'selectedClustering': fdata.selectedClustering}
# clusterings
for cID in fdata.clustering:
cStr = 'clust-%i-%i' % (dID, cID)
csett = {'method': fdata.methodIDs[cID],
'opts': fdata.clusteringOpts[cID],
'clusteringSelDims': fdata.clusteringSelDims[cID],
'infoExpanded': fdata.infoExpanded[cID]}
store[cStr] = csett
bindata[cStr] = fdata.clustering[cID]
# figures
sfigs = []
for figID in FigureStore.getFigures():
fig = FigureStore.get(figID)
fStr = 'fig-%i' % figID
d = dict(fig.__dict__)
splots = packSubplots(store, figID, fig.subplots)
d['subplots'] = splots
store[fStr] = d
sfigs.append(fStr)
store['figures'] = list(sfigs)
# other
store['current-data'] = DataStore.getCurrentIndex()
store['current-figure'] = FigureStore.getSelectedIndex()
# write out settings data
store.close()
# write out numeric data to binary file
np.savez(os.path.join(dir, binfile), **bindata)
def __init__(self, games, previous_length):
self.previous_length = previous_length
self.data_store = DataStore(previous_length)
self.games = [Attributes(game) for game in games]