attrs for context

Author: NicoRenaud

Diff for: DeepXplorer.py

@@ -33,8 +33,11 @@ class QIPythonWidget(RichJupyterWidget):
     Convenience class for a live IPython console widget. 
     We can replace the standard banner using the customBanner argument
     https://stackoverflow.com/questions/11513132/embedding-ipython-qt-console-in-a-pyqt-application
+    possible colors lightbg,linux,nocolor
     """
-    def __init__(self,customBanner=None,colors='linux',*args,**kwargs):
+
+
+    def __init__(self,customBanner=None,colors='lightbg',*args,**kwargs):
         if not customBanner is None: self.banner=customBanner
         super(QIPythonWidget, self).__init__(*args,**kwargs)
         self.kernel_manager = kernel_manager = QtInProcessKernelManager()
@@ -61,19 +64,29 @@ def printText(self,text):
         """ Prints some plain text to the console """
         self._append_plain_text(text)  
 
-    def executeCommand(self,command):
+    def executeCommand(self,command,hidden=False,interactive=False):
         """ Execute a command in the frame of the console widget """
-        self._execute(command,False)
+        self.execute(source=command,hidden=hidden,interactive=interactive)
 
     def import_value(self,HDF5Object):
         HDF5Object.emitDict.connect(self.get_value)
 
     @pyqtSlot(dict)
     def get_value(self,variableDict):
-        self.pushVariables(variableDict)
-        for k in variableDict:
-            self.printText(k + '\n')
-            self.executeCommand('print(%s)' %k)
+
+        # exectute a command
+        keys = list(variableDict.keys())
+        if keys[0].startswith('exec_cmd'):
+            for k,cmd in variableDict.items():
+                self._execute(cmd,False)
+
+        # push a variable in the console
+        # print it if the name doesnt start with _
+        else:
+            self.pushVariables(variableDict)
+            for k in variableDict:
+                if not k.startswith('_'):
+                    self.executeCommand('%s' %k)
 
 class DummyHDF5Group(dict):
     def __init__(self,dictionary, attrs ={}, name="DummyHDF5Group"):
@@ -321,21 +334,35 @@ def set_up_treeview(self, treeview):
         treeview.customContextMenuRequested.connect(functools.partial(self.context_menu, treeview))
 
     def context_menu(self, treeview, position):
+
         """Generate a right-click menu for the items"""
 
         # make sure tha there is only one item selected
         items = [self._index_to_item(index) for index in treeview.selectedIndexes()]
-        if len(items)>1:
+        if len(items)!=1:
             return
         item = items[0]
 
-        if len(item.name.split('/')) != 3:
-            return
+        try:
+            _type = self.root_item.data_file[item.name].attrs['type']
 
-        # no right click if no children
-        if not item._has_children:
+            if _type == 'molecule':
+                self._context_mol(item,treeview,position)
+
+            if _type == 'sparse_matrix':
+                self._context_sparse(item,treeview,position)
+
+            if _type == 'epoch':
+                self._context_epoch(item,treeview,position)
+
+            if _type == 'losses':
+                self._context_losses(item,treeview,position)
+
+        except:
             return
 
+    def _context_mol(self,item,treeview,position):
+
         menu = QtWidgets.QMenu()
         actions = {}
         list_operations = ['Load in PyMol','Load in VMD','PDB2SQL']
@@ -345,28 +372,152 @@ def context_menu(self, treeview, position):
         action = menu.exec_(treeview.viewport().mapToGlobal(position))
 
         if action == actions['Load in VMD']:
-            cplx_name = item.name.split('/')[1]
-            mol_name =  item.name.split('/')[2]
-            grp_name = '/' + cplx_name + '/' + mol_name
-            molgrp = self.root_item.data_file[grp_name]
+            _,cplx_name,mol_name = item.name.split('/')
+            molgrp = self.root_item.data_file[item.name]
             viztools.create3Ddata(mol_name,molgrp)
             viztools.launchVMD(mol_name,self.res)
 
         if action == actions['Load in PyMol']:
-            cplx_name = item.name.split('/')[1]
-            mol_name =  item.name.split('/')[2]
-            grp_name = '/' + cplx_name + '/' + mol_name
-            molgrp = self.root_item.data_file[grp_name]
+            _,cplx_name,mol_name = item.name.split('/')
+            molgrp = self.root_item.data_file[item.name]
             viztools.create3Ddata(mol_name,molgrp)
             viztools.launchPyMol(mol_name)
 
         if action == actions['PDB2SQL']:
-            cplx_name = item.name.split('/')[1]
-            mol_name =  item.name.split('/')[2]
-            grp_name = '/' + cplx_name + '/' + mol_name
-            molgrp = self.root_item.data_file[grp_name]
+            _,cplx_name,mol_name = item.name.split('/')
+            molgrp = self.root_item.data_file[item.name]
             db = pdb2sql(molgrp['complex'].value)
-            treeview.emitDict.emit({'_'+item.basename+'_pb2sql': db})
+            treeview.emitDict.emit({'sql_' + item.basename: db})
+
+    def _context_sparse(self,item,treeview,position):
+
+        menu = QtWidgets.QMenu()
+        actions = {}
+        list_operations = ['Load Matrix','Plot Histogram']
+
+        for operation in list_operations:
+            actions[operation] = menu.addAction(operation)
+        action = menu.exec_(treeview.viewport().mapToGlobal(position))
+        name = item.basename + '_' + item.name.split('/')[2]
+
+        if action == actions['Load Matrix']:
+
+            subgrp = item.data_file[item.name]
+            data_dict = {}
+            if not subgrp.attrs['sparse']:
+                data_dict[item.name] =  subgrp['value'].value 
+            else:
+                molgrp = item.data_file[item.parent.parent.parent.name]
+                grid = {}
+                lx = len(molgrp['grid_points/x'].value)
+                ly = len(molgrp['grid_points/y'].value)
+                lz = len(molgrp['grid_points/z'].value)
+                shape = (lx,ly,lz)
+                spg = sparse.FLANgrid(sparse=True,index=subgrp['index'].value,value=subgrp['value'].value,shape=shape)
+                data_dict[name] =  spg.to_dense()
+            treeview.emitDict.emit(data_dict)
+
+        if action == actions['Plot Histogram']:
+
+            value = item.data_file[item.name]['value'].value
+            data_dict = {'value':value}
+            treeview.emitDict.emit(data_dict)
+
+            cmd = "%matplotlib inline\nimport matplotlib.pyplot as plt\nplt.hist(value,25)\nplt.show()\n"            
+            data_dict = {'exec_cmd':cmd} 
+            treeview.emitDict.emit(data_dict)
+
+    def _context_epoch(self,item,treeview,position):
+
+        menu = QtWidgets.QMenu()
+        actions = {}
+        list_operations = ['Scatter Plot']
+
+        for operation in list_operations:
+            actions[operation] = menu.addAction(operation)
+        action = menu.exec_(treeview.viewport().mapToGlobal(position))
+        
+        if action == actions['Scatter Plot']:
+
+            
+            values = []
+            train_out = item.data_file[item.name+'/train/outputs'].value
+            train_tar = item.data_file[item.name+'/train/targets'].value
+            values.append([x for x in train_out])
+            values.append([x for x in train_tar])
+            
+
+            valid_out = item.data_file[item.name+'/valid/outputs'].value
+            valid_tar = item.data_file[item.name+'/valid/targets'].value
+            values.append([x for x in valid_tar])
+            values.append([x for x in valid_out])
+
+
+            test_out = item.data_file[item.name+'/test/outputs'].value
+            test_tar = item.data_file[item.name+'/test/targets'].value
+            values.append([x for x in test_tar])
+            values.append([x for x in test_out])
+            
+            
+            vmin = np.array([x for a in values for x in a]).min()
+            vmax = np.array([x for a in values for x in a]).max()
+            delta = vmax-vmin
+            values.append([vmax + 0.1*delta])
+            values.append([vmin - 0.1*delta])
+
+            data_dict = {'_values':values}
+            treeview.emitDict.emit(data_dict)
+
+            data_dict = {}
+            cmd  = "%matplotlib inline\nimport matplotlib.pyplot as plt\n"    
+            cmd += "fig,ax = plt.subplots()\n"
+            cmd += "ax.scatter(_values[0],_values[1],c='red',label='train')\n"
+            cmd += "ax.scatter(_values[2],_values[3],c='blue',label='valid')\n"
+            cmd += "ax.scatter(_values[4],_values[5],c='green',label='test')\n"
+            cmd += "legen = ax.legend(loc='upper left')\n"
+            cmd += "ax.set_xlabel('Targets')\n"
+            cmd += "ax.set_ylabel('Predictions')\n"
+            cmd += "ax.plot([_values[-2],_values[-1]],[_values[-2],_values[-1]])\n"
+            cmd += "plt.show()\n" 
+            data_dict['exec_cmd'] = cmd
+            treeview.emitDict.emit(data_dict)
+
+    def _context_losses(self,item,treeview,position):
+
+        menu = QtWidgets.QMenu()
+        actions = {}
+        list_operations = ['Plot Losses']
+
+        for operation in list_operations:
+            actions[operation] = menu.addAction(operation)
+        action = menu.exec_(treeview.viewport().mapToGlobal(position))
+        
+        if action == actions['Plot Losses']:
+
+            
+            values = []
+            test = item.data_file[item.name+'/test'].value
+            train = item.data_file[item.name+'/train'].value
+            valid = item.data_file[item.name+'/valid'].value
+            values.append([x for x in train])
+            values.append([x for x in valid])
+            values.append([x for x in test])
+            
+            data_dict = {'_values':values}
+            treeview.emitDict.emit(data_dict)
+
+            data_dict = {}
+            cmd  = "%matplotlib inline\nimport matplotlib.pyplot as plt\n"    
+            cmd += "fig,ax = plt.subplots()\n"
+            cmd += "plt.plot(_values[0],c='red',label='train')\n"
+            cmd += "plt.plot(_values[1],c='blue',label='valid')\n"
+            cmd += "plt.plot(_values[2],c='green',label='test')\n"
+            cmd += "legen = ax.legend(loc='upper right')\n"
+            cmd += "ax.set_xlabel('Epoch')\n"
+            cmd += "ax.set_ylabel('Losses')\n"
+            cmd += "plt.show()\n" 
+            data_dict['exec_cmd'] = cmd
+            treeview.emitDict.emit(data_dict)
 
 class HDF5TreeWidget(QtWidgets.QTreeView):
 
@@ -402,32 +553,13 @@ def doubleClicked(self,event):
 
         # get the current item
         items = [self.model._index_to_item(index) for index in self.selectedIndexes()]
-        if len(items)>1:
+        if len(items)!=1:
             return
         item=items[0]
-
-        # if we have a mapped feature
-        # we push the matrix
-        if item.parent.parent.basename == 'mapped_features':
-            subgrp = item.data_file[item.name]
-            data_dict = {}
-            name = item.name.replace('/','_')
-            if not subgrp.attrs['sparse']:
-                data_dict[item.name] =  subgrp['value'].value 
-            else:
-                molgrp = item.data_file[item.parent.parent.parent.name]
-                grid = {}
-                lx = len(molgrp['grid_points/x'].value)
-                ly = len(molgrp['grid_points/y'].value)
-                lz = len(molgrp['grid_points/z'].value)
-                shape = (lx,ly,lz)
-                spg = sparse.FLANgrid(sparse=True,index=subgrp['index'].value,value=subgrp['value'].value,shape=shape)
-                data_dict[name] =  spg.to_dense()
-            self.emitDict.emit(data_dict)
-
+        
         # if the item has no children we push the raw data
-        elif not item._has_children:
-            name = item.name.replace('/','_')
+        if not item._has_children:
+            name = item.basename + '_' + item.name.split('/')[2]
             self.emitDict.emit({name: item.data_file[item.name].value})
 
         # or we skip

Diff for: README.md

@@ -1,6 +1,6 @@
 # DeepXplorer
 
-Data browser for the HDF5 files created by deeprank. The GUI allows to easily explore file structure, view, the target values and visualize the mapped features in either VMD or PyMol. In addition an the data can be loaded in an embedded iPython console for manipulation and analysis. 
+Data browser for the HDF5 files created by deeprank. The GUI allows to easily explore file structure, view, the target values and visualize the mapped features in either VMD or PyMol. In addition  the data can be loaded in an embedded iPython console for manipulation and analysis. 
 
 ![alt-text](./deepxplorer.png)