Updated PCA organization

Entropy/Entropy.py

@@ -716,7 +716,9 @@ def plotCCpca(self,index=None,states:list=None,ax=None,**kwargs):
 
         if 'color' not in kwargs:kwargs['color']='black'
         ax.plot(self.CCpca[index],**kwargs)
+        ax.xaxis.set_major_locator(plt.MaxNLocator(30,integer=True))
         ax.xaxis.set_major_formatter(self._axis_formatter(index))
+
         ax.tick_params(axis='x', labelrotation=90)
         ax.set_ylabel('C.C')
 
@@ -786,6 +788,8 @@ def plotCC(self,index=None,ax=None,CCsum:bool=True,**kwargs):
 
         ax.xaxis.set_major_formatter(self._axis_formatter(index))
         ax.yaxis.set_major_formatter(self._axis_formatter(index))
+        ax.xaxis.set_major_locator(plt.MaxNLocator(30,integer=True))
+        ax.yaxis.set_major_locator(plt.MaxNLocator(30,integer=True))
         ax.tick_params(axis='x', labelrotation=90)
         # ax.set_aspect('auto')
         if CCsum:

IO/bin_IO.py

@@ -561,6 +561,19 @@ def find_file(filename,directory:str='',f=None):
 
 
 
+# def write_PCA(f: typing.BinaryIO, pca: pyDR.PCA):
+#     f.write(b'OBJECT:PCA\n')
+#     write_MolSelect(pca.select)
+
+#     pca_flds=[]
+
+#     for k in pca_flds:
+#         f.write(bytes('{0}\n'.format(k),'utf-8'))
+#         np.save(f,getattr(pca,k),allow_pickle=False)
+
+
+#     f.write(b'END:OBJECT\n')
+
 
 
 

PCA/CombinedPCA.py

@@ -0,0 +1,197 @@
+# -*- coding: utf-8 -*-
+
+
+from . import PCA
+from ..Selection.MolSys import Trajectory
+import numpy as np
+import matplotlib.pyplot as plt
+from matplotlib.colors import ListedColormap
+from copy import copy
+
+class CombinedPCA(PCA):
+    def __init__(self,*pcas):
+        super().__init__(pcas[0].select)
+        # self.__dict__=copy(pcas[0].__dict__)
+        self._pos=np.concatenate([pca0.pos for pca0 in pcas],axis=0)
+        self.pcas=[copy(pca) for pca in pcas]
+
+        self._traj=Traj(self)
+
+        for pca,Range in zip(self.pcas,self.ranges):
+             pca._PC=self.PC
+             pca._pcamp=self.PCamp[:,Range[0]:Range[1]]
+             pca._lambda=self.Lambda
+
+        self.traj[0]
+
+
+    @property
+    def traj(self):
+        return self._traj
+
+    @property
+    def n_trajs(self):
+        return len(self.trajs)
+
+    @property
+    def trajs(self):
+        return [pca.traj for pca in self.pcas]
+
+    @property
+    def lengths(self):
+        return np.array([len(traj) for traj in self.trajs],dtype=int)
+
+    @property
+    def ranges(self):
+        return [(self.lengths[:n].sum(),self.lengths[:n+1].sum()) for n in range(self.n_trajs)]
+
+    def hist_by_traj(self,nmax:int=3,cmap='Reds',cmap0='jet',**kwargs):
+        fig,ax=plt.subplots(nmax,self.n_trajs)
+        if nmax==1:ax=np.array([ax])
+
+        cm=plt.get_cmap(cmap)
+        colors=np.array([cm(k) for k in range(256)])
+        colors[:,-1]=np.linspace(0,1,257)[1:]**.25
+        cm=ListedColormap(colors)
+
+        for q,ax0 in enumerate(ax.T):
+            for ax00,n0,n1 in zip(ax0,range(nmax),range(1,nmax+1)):
+                self.Hist.plot(n0,n1,ax=ax00,cmap=cmap0,**kwargs)
+                index=np.zeros(len(self.traj),dtype=bool)
+                index[self.ranges[q][0]:self.ranges[q][1]]=True
+                self.Hist.plot(n0,n1,ax=ax00,cmap=cm,index=index)
+
+        fig.tight_layout()
+
+        return fig
+
+    def hist2struct(self,nmax:int=3,from_traj:bool=True,select_str:str='protein',
+                    ref_struct:bool=False,cmap='Reds',cmap0='jet',cmap_ch='gist_rainbow',n_colors=10,**kwargs):
+        fig,ax=plt.subplots(nmax,self.n_trajs)
+
+        cm=plt.get_cmap(cmap)
+        colors=np.array([cm(k) for k in range(255)])
+        colors[:,-1]=np.linspace(0,1,255)**.25
+        cm=ListedColormap(colors)
+
+        for q,ax0 in enumerate(ax.T):
+            xlims=[]
+            ylims=[]
+            for ax00,n0,n1 in zip(ax0,range(nmax),range(1,nmax+1)):
+                self.Hist.plot(n0,n1,ax=ax00,cmap=cmap0,**kwargs)
+                ax00.set_xlim(ax00.get_xlim())
+                ax00.set_ylim(ax00.get_ylim())
+                xlims.append(ax00.get_xlim())
+                ylims.append(ax00.get_ylim())
+            self.pcas[q].Hist.hist2struct(nmax=nmax,from_traj=from_traj,select_str=select_str,
+                                        ref_struct=ref_struct,ax=ax0.tolist(),
+                                        cmap=cm,cmap_ch=cmap_ch,n_colors=n_colors,**kwargs)
+            for ax00,xlim,ylim in zip(ax0,xlims,ylims):
+                ax00.set_xlim(xlim)
+                ax00.set_ylim(ylim)
+
+        return fig
+
+    def hist_t_depend(self,nmax:int=3,cmap='jet',cmap0='Greys',step:int=20,**kwargs):
+        fig,ax=plt.subplots(nmax,self.n_trajs)
+        if nmax==1:ax=np.array([ax])
+
+        if isinstance(cmap,str):cmap=plt.get_cmap(cmap)
+
+
+        for q,ax0 in enumerate(ax.T):
+            for ax00,n0,n1 in zip(ax0,range(nmax),range(1,nmax+1)):
+                self.Hist.plot(n0,n1,ax=ax00,cmap=cmap0,**kwargs)
+                index=np.zeros(len(self.traj),dtype=bool)
+                index[self.ranges[q][0]:self.ranges[q][1]:step]=True
+                cmap=cmap.resampled(index.sum())
+                c=cmap(np.arange(index.sum()))
+                ax00.scatter(self.PCamp[n0][index],self.PCamp[n1][index],s=1,c=c)
+
+        fig.tight_layout()
+
+        return fig
+
+
+
+
+
+    @property
+    def filenames(self):
+        return [pca.uni.filename for pca in self.pcas]
+
+
+
+
+
+
+
+class Traj():
+    def __init__(self,combinePCA):
+        self.cPCA=combinePCA
+        self._index=0
+        self._traj_index=0
+        self.mda_traj=self.trajs[0].mda_traj
+
+    @property
+    def trajs(self):
+        return self.cPCA.trajs
+
+    @property
+    def pcas(self):
+        return self.cPCA.pcas
+
+    @property
+    def lengths(self):
+        return self.cPCA.lengths
+
+    def __getitem__(self,i):
+        # q=np.argmax(i<np.cumsum(np.concatenate([[0],self.cPCA.lengths])))
+        q=np.argmax(i<np.cumsum(self.cPCA.lengths))
+        pca=self.pcas[q]
+        self.cPCA._uni=pca.uni
+        self.cPCA._atoms=pca.atoms
+        self.mda_traj=pca.traj.mda_traj
+        self._traj_index=q
+        self._index=i
+        return pca.traj[(i%len(self))-self.lengths[:q].sum()]
+
+    @property
+    def traj_index(self):
+        return self._traj_index
+
+    @property
+    def index(self):
+        return self._index
+
+
+    def __len__(self):
+        return self.tf
+
+    @property
+    def t0(self):
+        return 0
+
+    @t0.setter
+    def t0(self,t0):
+        pass
+
+    @property
+    def tf(self):
+        return np.sum([len(traj) for traj in self.cPCA.trajs])
+
+    @tf.setter
+    def tf(self,tf):
+        pass
+
+    @property
+    def step(self):
+        return self.cPCA.trajs[0].step
+
+    @step.setter
+    def step(self,step):
+        pass
+
+    @property
+    def frame(self):
+        return (self.mda_traj.frame-self.trajs[self._traj_index].t0)//self.step