python based data analysis

data/tools/support/maximum.py

+import numpy as np
+
+
+def maximum(x):
+    size = x.shape
+    res = np.zeros(size[0])
+
+    for time_step in range(size[0]):
+        res[time_step] = max(res[time_step], max(x[time_step, :]))
+
+    return res

data/tools/support/maximum.py~

+import numpy as np
+
+def maximum(x) :
+    size = x.shape 
+    res = np.zeros(size[0])
+
+    for time_step in range(size[0]) :
+        for vertex in range(size[1]) :
+            res[time_step] = max(res[time_step], max(x[time_step, vertex]))
+
+    return res
\ No newline at end of file

data/tools/support/negativeStarts.cpp

-#include <Rcpp.h>
-
-// [[Rcpp::export]]
-Rcpp::NumericVector negativeStarts(Rcpp::NumericVector const &x) {
-  std::vector<size_t> starts(0);
-  bool prev = false;
-  for (size_t i = 0; i < x.size(); ++i) {
-    if (prev && !x[i])
-      starts.push_back(i+1); // R indices start at 1
-    prev = x[i];
-  }
-  return Rcpp::NumericVector(starts.begin(), starts.end());
-}

data/tools/support/norm.py

+import numpy as np
+
+
+def norm(x):
+    size = x.shape
+    res = np.zeros((size[0], size[1]))
+
+    for time_step in range(size[0]):
+        for vertex in range(size[1]):
+            res[time_step, vertex] = np.linalg.norm(x[time_step, vertex, :])
+
+    return res

data/tools/support/norm.py~

+import numpy as np
+
+def norm(x) :
+    size = x.shape 
+    res = np.zeros(size[0], size[1])
+
+    for time_step in range(size[0]) :
+        for vertex in range(size[1]) :
+            res[time_step, vertex] = np.linalg.norm(x[time_step, vertex, :])
+
+    return res 

data/tools/support/positiveStarts.cpp

-#include <Rcpp.h>
-
-// [[Rcpp::export]]
-Rcpp::NumericVector positiveStarts(Rcpp::NumericVector const &x) {
-  std::vector<size_t> starts(0);
-  bool prev = false;
-  for (size_t i = 0; i < x.size(); ++i) {
-    if (!prev && x[i])
-      starts.push_back(i+1); // R indices start at 1
-    prev = x[i];
-  }
-  return Rcpp::NumericVector(starts.begin(), starts.end());
-}
-

data/tools/support/slip_beginnings.py

+import array as ar
+
+
+def slip_beginnings(x):
+    # returns indicies i for which x[i-1]=0 and x[i]>0
+    starts = ar.array('i', (0 for i in range(len(x))))
+    prev = False
+
+    length = 0
+    for i in range(len(x)):
+        if (not prev and x[i]):
+            starts[length] = i
+            length += 1
+        prev = x[i]
+    return starts[:length]

data/tools/support/slip_beginnings.py~

+import numpy as np
+
+def slip_beginnings(x) :
+    # returns indicies i for which x[i-1]=0 and x[i]>0  
+    starts = []
+    prev = False
+    for i in range(len(x)) :
+        if (not prev and x[i]) :
+            starts.append(i)
+            prev = x[i]
+    return starts

data/tools/support/slip_endings.py

+import array as ar
+
+
+def slip_endings(x):
+    # returns indicies i for which x[i-1]>0 and x[i]=0
+    ends = ar.array('i', (0 for i in range(len(x))))
+    prev = False
+
+    length = 0
+    for i in range(len(x)):
+        if (prev and not x[i]):
+            ends[length] = i
+            length += 1
+        prev = x[i]
+
+    return ends[:length]

data/tools/support/slip_endings.py~

+import numpy as np
+
+def slip_endings(x) :
+    # returns indicies i for which x[i-1]>0 and x[i]=0
+    starts = np.array([])
+    prev = False
+    for i in range(len(x)) :
+        if (prev and not x[i]) :
+            starts.append(i)
+            prev = x[i]
+    return starts
\ No newline at end of file

data/tools/support/trapezoidal.cpp

-#include <Rcpp.h>
-
-// [[Rcpp::export]]
-double trapezoidal(Rcpp::NumericVector const &x, Rcpp::NumericVector const &y) {
-  double ret = 0;
-  for (size_t i = 1; i < x.size(); ++i)
-    ret += (x[i] - x[i - 1]) * (y[i] + y[i - 1]) / 2;
-  return ret;
-}

data/tools/support/trapezoidal.py

+# x: vector of domain values
+# y: vector of function values f(x)
+def trapezoidal(x, y) :
+    # returns integral of (x, y = f(x)) calculated with trapezoidal rule 
+	res = 0.0
+	for i in range(1, len(x)) :
+		res += (x[i] - x[i - 1]) * (y[i] + y[i - 1]) / 2.0
+	return res
\ No newline at end of file

data/tools/support/trapezoidal.py~

+# x: vector of domain values
+# y: vector of function values f(x)
+def trapezoidal(x, y) :
+	# returns integral of (x, y = f(x)) calculated with trapezoidal rule 
+	res = 0.0
+	for i in range(1, len(x)) :
+		res += (x[i] - x[i - 1]) * (y[i] + y[i - 1]) / 2.0
+	return res
\ No newline at end of file

debugging.m

+% householder reflection
+
+v0 = [1; 0];
+v0 = v0./norm(v0);
+
+v1 = [1; -1];
+v1 = v1./norm(v1);
+
+v = 1/2.*(v1-v0);
+vnorm = norm(v);
+
+H = eye(2) - 2/(vnorm^2).*(v*v')
 
 A = [...
 ];