Add files via upload

Fisherinfor/load_statistics.py

@@ -0,0 +1,228 @@
+import numpy as np
+
+#-------------------------------------------------------
+def load_scale_pk_fiducial( root_path, nsims_fiducial, Imax ):
+    # initialize
+    statistic_fiducial = []
+
+    # Load the scalePKHF from fiducial
+    for i in range(nsims_fiducial): 
+        peaks   = np.load(root_path+'fiducial/peaks/'+str(i)+'/z=0/peaks.npz')
+        pkhf    = np.ravel( peaks['scalePKHF'][0:Imax,:] )
+        statistic_fiducial.append( pkhf )
+
+    return np.asarray(statistic_fiducial)
+
+#-------------------------------------------------------
+def load_scale_vly_fiducial( root_path, nsims_fiducial, Imax ):
+    # initialize
+    statistic_fiducial = []
+
+    # Load the scalePKHF and scaleVLYDF from fiducial
+    for i in range(nsims_fiducial): 
+        valleys = np.load(root_path+'fiducial/valleys/'+str(i)+'/z=0/valleys.npz')
+        vlydf   = np.ravel( valleys['scaleVLYDF'][0:Imax,:] )
+        statistic_fiducial.append( vlydf )
+
+    return np.asarray(statistic_fiducial)
+
+#-------------------------------------------------------
+def load_scale_extrema_fiducial( root_path, nsims_fiducial, Imax ):
+    # initialize
+    statistic_fiducial = []
+
+    # Load the scalePKHF and scaleVLYDF from fiducial
+    for i in range(nsims_fiducial): 
+        valleys = np.load(root_path+'fiducial/valleys/'+str(i)+'/z=0/valleys.npz')
+        peaks   = np.load(root_path+'fiducial/peaks/'+str(i)+'/z=0/peaks.npz')
+
+        vlydf   = np.ravel( valleys['scaleVLYDF'][0:Imax,:] )
+        pkhf    = np.ravel( peaks['scalePKHF'][0:Imax,:] )
+        stat    = np.concatenate((vlydf, pkhf)) 
+        statistic_fiducial.append( stat )
+
+
+    return np.asarray(statistic_fiducial)
+
+#-------------------------------------------------------
+def load_fps_fiducial( root_path, nsims_fiducial, kmax ):
+    # initialize
+    statistic_fiducial = []
+
+    # Load the scalePKHF and scaleVLYDF from fiducial
+    for i in range(nsims_fiducial): 
+        fps  = np.load(root_path+'fiducial/fps/'+str(i)+'/z=0/fps.npz')
+        if ( i == 0 ):
+            k = fps['k']
+            kk = k[k<kmax]
+
+        Pk = fps['Pk']  
+        Pkk = Pk[0:len(kk)-1]
+        statistic_fiducial.append( Pkk )
+
+    return np.asarray(statistic_fiducial)
+
+#-------------------------------------------------------
+def load_all_fiducial( root_path, nsims_fiducial, Imax, kmax ):
+    # initialize
+    statistic_fiducial = []
+
+    # Load the scalePKHF and scaleVLYDF from fiducial
+    for i in range(nsims_fiducial): 
+        valleys = np.load(root_path+'fiducial/valleys/'+str(i)+'/z=0/valleys.npz')
+        peaks   = np.load(root_path+'fiducial/peaks/'+str(i)+'/z=0/peaks.npz')
+        fps     = np.load(root_path+'fiducial/fps/'+str(i)+'/z=0/fps.npz')
+
+        if ( i == 0 ):
+            k = fps['k']
+            kk = k[k<kmax]
+
+        vlydf   = np.ravel( valleys['scaleVLYDF'][0:Imax,:] )
+        pkhf    = np.ravel( peaks['scalePKHF'][0:Imax,:] )
+        Pk      = fps['Pk'][0:len(kk)-1]  
+        stat    = np.concatenate( (vlydf, pkhf, Pk) )
+        statistic_fiducial.append( stat )
+
+    return np.asarray(statistic_fiducial)
+
+#-------------------------------------------------------
+def load_scale_pk_derivatives(root_path, paras, nsims_derivatives, Imax):
+    # initialize
+    statistics = {}  # dictionary
+    for para in paras:
+        statistics[para] = None
+
+    # Load the scaleVLYDF and scalePKHF from EQ, LS, OR_LSS, ...
+    for para in paras:
+        stat_p   = []
+        stat_m   = []
+        for i in range(nsims_derivatives):   
+            peaks_p   = np.load(root_path+para+'_p/peaks/'+str(i)+'/z=0/peaks.npz')
+            peaks_m   = np.load(root_path+para+'_m/peaks/'+str(i)+'/z=0/peaks.npz')
+            pkhf_p    = np.ravel( peaks_p['scalePKHF'][0:Imax,:] )
+            pkhf_m    = np.ravel( peaks_m['scalePKHF'][0:Imax,:] )
+
+            stat_p.append( pkhf_p )
+            stat_m.append( pkhf_m )
+
+        statistics[para] = [stat_p, stat_m]
+
+    return statistics
+
+#-------------------------------------------------------
+def load_scale_vly_derivatives(root_path, paras, nsims_derivatives, Imax):
+    # initialize
+    statistics = {}  # dictionary
+    for para in paras:
+        statistics[para] = None
+
+    # Load the scaleVLYDF and scalePKHF from EQ, LS, OR_LSS, ...
+    for para in paras:
+        stat_p   = []
+        stat_m   = []
+        for i in range(nsims_derivatives):   
+            valleys_p = np.load(root_path+para+'_p/valleys/'+str(i)+'/z=0/valleys.npz')
+            valleys_m = np.load(root_path+para+'_m/valleys/'+str(i)+'/z=0/valleys.npz')
+            vlydf_p   = np.ravel( valleys_p['scaleVLYDF'][0:Imax,:] )
+            vlydf_m   = np.ravel( valleys_m['scaleVLYDF'][0:Imax,:] )
+
+
+            stat_p.append( vlydf_p  )
+            stat_m.append( vlydf_m  )
+
+        statistics[para] = [stat_p, stat_m]
+
+    return statistics
+
+#-------------------------------------------------------
+def load_scale_extrema_derivatives(root_path, paras, nsims_derivatives, Imax):
+    # initialize
+    statistics = {}  # dictionary
+    for para in paras:
+        statistics[para] = None
+
+    # Load the scaleVLYDF and scalePKHF from EQ, LS, OR_LSS, ...
+    for para in paras:
+        stat_p   = []
+        stat_m   = []
+        for i in range(nsims_derivatives):   
+            valleys_p = np.load(root_path+para+'_p/valleys/'+str(i)+'/z=0/valleys.npz')
+            peaks_p   = np.load(root_path+para+'_p/peaks/'+str(i)+'/z=0/peaks.npz')
+            valleys_m = np.load(root_path+para+'_m/valleys/'+str(i)+'/z=0/valleys.npz')
+            peaks_m   = np.load(root_path+para+'_m/peaks/'+str(i)+'/z=0/peaks.npz')
+
+            vlydf_p   = np.ravel( valleys_p['scaleVLYDF'][0:Imax,:] )
+            pkhf_p    = np.ravel( peaks_p['scalePKHF'][0:Imax,:] )
+            vlydf_m   = np.ravel( valleys_m['scaleVLYDF'][0:Imax,:] )
+            pkhf_m    = np.ravel( peaks_m['scalePKHF'][0:Imax,:] )
+
+            stat_p.append( np.concatenate((vlydf_p, pkhf_p))  )
+            stat_m.append( np.concatenate((vlydf_m, pkhf_m))  )
+
+        statistics[para] = [stat_p, stat_m]
+
+    return statistics
+
+#-------------------------------------------------------
+def load_fps_derivatives(root_path, paras, nsims_derivatives, kmax):
+    # initialize
+    statistics = {}  # dictionary
+    for para in paras:
+        statistics[para] = None
+
+    # Load the scaleVLYDF and scalePKHF from EQ, LS, OR_LSS, ...
+    for para in paras:
+        stat_p   = []
+        stat_m   = []
+        for i in range(nsims_derivatives): 
+            fps_p = np.load(root_path+para+'_p/fps/'+str(i)+'/z=0/fps.npz')
+            fps_m = np.load(root_path+para+'_m/fps/'+str(i)+'/z=0/fps.npz')  
+            if ( i == 0 ):
+                k  = fps_p['k']
+                kk = k[k<kmax]
+            Pk_p = fps_p['Pk'][0:len(kk)-1]
+            Pk_m = fps_m['Pk'][0:len(kk)-1]
+
+
+            stat_p.append( Pk_p  )
+            stat_m.append( Pk_m  )
+
+        statistics[para] = [stat_p, stat_m]
+
+    return statistics
+
+#-------------------------------------------------------
+def load_all_derivatives(root_path, paras, nsims_derivatives, Imax, kmax):
+    # initialize
+    statistics = {}  # dictionary
+    for para in paras:
+        statistics[para] = None
+
+    # Load the scaleVLYDF and scalePKHF from EQ, LS, OR_LSS, ...
+    for para in paras:
+        stat_p   = []
+        stat_m   = []
+        for i in range(nsims_derivatives):   
+            valleys_p = np.load(root_path+para+'_p/valleys/'+str(i)+'/z=0/valleys.npz')
+            peaks_p   = np.load(root_path+para+'_p/peaks/'+str(i)+'/z=0/peaks.npz')
+            valleys_m = np.load(root_path+para+'_m/valleys/'+str(i)+'/z=0/valleys.npz')
+            peaks_m   = np.load(root_path+para+'_m/peaks/'+str(i)+'/z=0/peaks.npz')
+            fps_p     = np.load(root_path+para+'_p/fps/'+str(i)+'/z=0/fps.npz')
+            fps_m     = np.load(root_path+para+'_m/fps/'+str(i)+'/z=0/fps.npz') 
+            if ( i == 0 ):
+                k  = fps_p['k']
+                kk = k[k<kmax]
+
+            vlydf_p   = np.ravel( valleys_p['scaleVLYDF'][0:Imax,:] )
+            pkhf_p    = np.ravel( peaks_p['scalePKHF'][0:Imax,:] )
+            vlydf_m   = np.ravel( valleys_m['scaleVLYDF'][0:Imax,:] )
+            pkhf_m    = np.ravel( peaks_m['scalePKHF'][0:Imax,:] )
+            Pk_p      = fps_p['Pk'][0:len(kk)-1]
+            Pk_m      = fps_m['Pk'][0:len(kk)-1]
+
+            stat_p.append( np.concatenate((vlydf_p, pkhf_p, Pk_p))  )
+            stat_m.append( np.concatenate((vlydf_m, pkhf_m, Pk_m))  )
+
+        statistics[para] = [stat_p, stat_m]
+
+    return statistics