diff --git a/scripts/loaddata.py b/scripts/example_loaddata.py
similarity index 100%
rename from scripts/loaddata.py
rename to scripts/example_loaddata.py
diff --git a/scripts/test.py b/scripts/example_test.py
similarity index 100%
rename from scripts/test.py
rename to scripts/example_test.py
diff --git a/scripts/isomap_large.py b/scripts/isomap_large.py
index 4488408d23174d1b8bc0e12f9cebeafb769acf3e..602b2249a93c26a38d6596a9601d6a30bd709987 100644
--- a/scripts/isomap_large.py
+++ b/scripts/isomap_large.py
@@ -4,6 +4,9 @@
Created on Tue Jun 12 15:40:44 2018
@author: xiayu
+
+the first index contains a num of 12048 points, whitch is too large for PC's
+memory capacity. run this on a server.
"""
import numpy as np
@@ -16,10 +19,8 @@ with np.load('../data/stddata.npz') as d:
with np.load('../data/sifteredindex.npz') as d:
data = d['data']
-X_ind = data[6]['index']
+X_ind = data[0]['index']
iso = manifold.Isomap(30, n_components=12)
phi_s = getmapped(phi, X_ind, iso)
-iso2 = manifold.Isomap(30, n_components=2)
-phi_s2 = getmapped(phi, X_ind, iso2)
-#np.savez_compressed('../data/sifteredisomappeddatalarge',data=phi_s)
+np.savez_compressed('../data/sifteredisomappeddatalarge',data=phi_s)
diff --git a/scripts/isomappedtomat.py b/scripts/isomappedtomat.py
index 84a65b1e68274df2b36534f3c78f324ed42ef6aa..0d8af4c3dfc4fa9312040a08dee772b812d1bdd3 100644
--- a/scripts/isomappedtomat.py
+++ b/scripts/isomappedtomat.py
@@ -6,10 +6,6 @@ Created on Tue Jun 12 16:38:07 2018
@author: xiayu
"""
-from scipy.io import savemat
-import numpy as np
-
-with np.load('../data/stddata.npz') as d:
- savemat('../data/stddata.mat',d)
-
+from pkgs.convertor import npz2mat
+npz2mat('../data/isomappeddata.npz','../data/isomappeddata.mat')
diff --git a/scripts/pkgs/__pycache__/__init__.cpython-36.pyc b/scripts/pkgs/__pycache__/__init__.cpython-36.pyc
index fd86a4153888db0802bde6b1e5c331f979d74a1b..f8392d92ce392d7725e41bba55b436acc1fe63b0 100644
Binary files a/scripts/pkgs/__pycache__/__init__.cpython-36.pyc and b/scripts/pkgs/__pycache__/__init__.cpython-36.pyc differ
diff --git a/scripts/pkgs/__pycache__/convertor.cpython-36.pyc b/scripts/pkgs/__pycache__/convertor.cpython-36.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..fa5d399cb714dae0c8278a003736c664b469a95e
Binary files /dev/null and b/scripts/pkgs/__pycache__/convertor.cpython-36.pyc differ
diff --git a/scripts/pkgs/convertor.py b/scripts/pkgs/convertor.py
new file mode 100644
index 0000000000000000000000000000000000000000..3aec6f85bdf18f36b06dcfcae628c8a4d87f7b5c
--- /dev/null
+++ b/scripts/pkgs/convertor.py
@@ -0,0 +1,16 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Created on Thu Jun 14 11:30:48 2018
+
+@author: xiayu
+"""
+
+from scipy.io import savemat
+from numpy import load
+
+def npz2mat(npz, mat):
+ with load(npz) as d:
+ savemat(mat, d)
+ pass
+
diff --git a/scripts/pkgs/mapping.py b/scripts/pkgs/mapping.py
index 6c26498fc144ec30dcda0d9388c2870b4a3bf606..9815b547f770610b5c27ae485d876ab423b34fbf 100644
--- a/scripts/pkgs/mapping.py
+++ b/scripts/pkgs/mapping.py
@@ -9,6 +9,15 @@ Created on Tue Jun 12 11:30:26 2018
import numpy as np
def getmapped(phi, X_ind, mapping):
+ '''
+ :para phi: full set of points.
+ :para X_ind: index of points which are seclected as sample for mapping.
+ normally we don't use full set of points for mapping analysis.
+ :para mapping: mapping object from sklearn.manifold.
+ :return: full set of mapped points. selected points are directly mapped,
+ while the rest points are mapped by a out-of-sample algorithm
+ provided by the mapping object.
+ '''
X = phi[X_ind]
X_s = mapping.fit_transform(X)
diff --git a/scripts/pkgs/sifter.py b/scripts/pkgs/sifter.py
index f457633c19d982e8046b33c986074257c607014b..236bbec3609bc9fa8a836c1a4f77b221e2d8022c 100644
--- a/scripts/pkgs/sifter.py
+++ b/scripts/pkgs/sifter.py
@@ -10,6 +10,11 @@ import numpy as np
from sklearn.neighbors import NearestNeighbors
def radius_sifter(X,radius):
+ """
+ sifter X with radius, so that every two points remain is
+ with distance > radius
+ """
+
neigh = NearestNeighbors(radius=radius)
indices = np.empty(X.shape[0],dtype=int)
ind_rest = np.arange(X.shape[0])
diff --git a/scripts/sifterindex.py b/scripts/sifterindex.py
deleted file mode 100644
index 1abe0a0d0ed1da3a7f66f5a350a5638c81624f1c..0000000000000000000000000000000000000000
--- a/scripts/sifterindex.py
+++ /dev/null
@@ -1,21 +0,0 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-"""
-Created on Tue Jun 12 11:57:35 2018
-
-@author: xiayu
-"""
-
-import numpy as np
-
-with np.load('../data/siftereddata.npz') as d:
- data = d['data']
-
-dt = np.dtype([('index', 'O')])
-a = np.array([],dtype=dt)
-
-for i in range(1,8):
- b = np.array([(data[i]['index'],)],dtype=dt)
- a = np.append(a,b)
-
-np.savez_compressed('../data/sifteredindex',data=a)
\ No newline at end of file
diff --git a/scripts/convertdata.py b/scripts/step1_convert_data.py
similarity index 100%
rename from scripts/convertdata.py
rename to scripts/step1_convert_data.py
diff --git a/scripts/scaleddata.py b/scripts/step2_scale_data.py
similarity index 100%
rename from scripts/scaleddata.py
rename to scripts/step2_scale_data.py
diff --git a/scripts/get_sifter.py b/scripts/step3_sift_data.py
similarity index 59%
rename from scripts/get_sifter.py
rename to scripts/step3_sift_data.py
index 5fb39192357486f7c29ac2cfa632477ac035b494..ac622e6a4379856a4135e4f5ee312fc0b50034ff 100644
--- a/scripts/get_sifter.py
+++ b/scripts/step3_sift_data.py
@@ -4,6 +4,8 @@
Created on Sun Jan 7 17:08:10 2018
@author: xiayu
+
+only index is required for the sake of space saving
"""
from pkgs.sifter import radius_sifter
@@ -25,4 +27,16 @@ for r in radius_list:
b = np.array([(phi,ind)],dtype=dt)
a = np.append(a,b)
-np.savez_compressed('../data/siftereddata',data=a)
\ No newline at end of file
+np.savez_compressed('../data/siftereddata',data=a)
+
+with np.load('../data/siftereddata.npz') as d:
+ data = d['data']
+
+dt = np.dtype([('index', 'O')])
+a = np.array([],dtype=dt)
+
+for i in range(1,8):
+ b = np.array([(data[i]['index'],)],dtype=dt)
+ a = np.append(a,b)
+
+np.savez_compressed('../data/sifteredindex',data=a)
\ No newline at end of file