Extend my __init__ method by an extra parameter Tol

[duyduy]

I want to extend my __init__ method by an extra parameter Tol (tolerance) which serves as a threshold for converting a given matrix to a sparse matrix. All elements of the original matrix which have an absolute value less than Tol should be considered as zeros. Give Tol a default value 0

this is my code

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53

import numpy as np     class SparseMatrix:        def __init__(self,matrix):         self.matrix=np.asarray(matrix)         self.intern_represent = 'CSR'         self.intern_represent = 'CSC'           @property       def values (self):         return self.matrix [np.nonzero(self.matrix)]                        def __repr__(self):         return (' A= '+ str(self.values) +                 '\nIA= ' + str(self.IA) +                 '\nJA= '+ str(self.JA))        @property     def IA (self):         IA =[]         IA.append (0)         for i in range (1,np.shape(self.matrix)[0]+1):             num=IA [i-1] + np.count_nonzero(self.matrix [i-1])             IA.append(num)         IA=np.asarray (IA)         return IA                @property     def JA (self):         index=np.nonzero(self.matrix)         return index [1]             @property       def number_of_nonzero(self):         return len (self.values)                 def change (self,i,j,new):         self.matrix [i][j]=new         return new       def ConvertCSC(self):         return np.transpose(self.matrix)         def CSReqaulCSC(self):       if (self.matrix==np.transpose(self.matrix)).all():           return True       else:           return False       all

[Yoriz]

1	def __init__(self,matrix, tol=0):

[duyduy]

That part I got, but how can i use it when i want to run the code. Lets say i have this matrix

1 2 3 4 5 6

a = SparseMatrix([[1,3,0,0,0],                   [0,2,0,3,0],                   [0,1,1,0,0],                   [3,0,0,0,2],                   [0,1,0,0,0]])

and I want my tolerance to filter all my numbers higher than 1. Do I change in "def value" or my "IA"?
thx for replying!

[Yoriz]

1 2 3 4 5 6 7

class SparseMatrix:         def __init__(self,matrix, tol=0):         self.matrix=np.asarray(matrix)         self.intern_represent = 'CSR'         self.intern_represent = 'CSC'         self.tol = tol

then use self.tol anywhere you need to use it in the class

---

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

class SparseMatrix:          def __init__(self,matrix, tol=0):         self.matrix=np.asarray(matrix)         self.intern_represent = 'CSR'         self.intern_represent = 'CSC'         self.tol = tol       def show_tol_value(self):         print(self.tol)     use_default_tol = SparseMatrix(matrix) use_default_tol.show_tol_value() change_default_tol = SparseMatrix(matrix, 1) change_default_tol.show_tol_value()

Output:
0
1

[duyduy]

"All elements of the original matrix which have an absolute value less than Tol should be considered as zeros"

what do you think he(my teacher) means by that quote?

[Yoriz]

If a value is greater then tol use the value, but if a value is less then tol, 0 should be used instead of the value.

[jefsummers]

Missing the issue of negative numbers. If a value's absolute value is less than tol, change the value to zero.

[DeaD_EyE]

If you use numpy, it's easier:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

In [1]: import numpy as np                                                                                                              In [2]: np.random.randint(1, 100, (3,3))                                                                                              Out[2]: array([[58, 13,  8],        [16, 41, 76],        [44, 22, 32]])   In [3]: np.random.randint(1, 10, (3,3))                                                                                               Out[3]: array([[9, 9, 4],        [6, 7, 6],        [5, 9, 9]])   In [4]: matrix = np.random.randint(1, 10, (3,3))                                                                                        In [5]: matrix                                                                                                                        Out[5]: array([[9, 6, 3],        [6, 7, 4],        [7, 4, 2]])   In [6]: matrix[matrix < 5]                                                                                                            Out[6]: array([3, 4, 4, 2])   In [7]: matrix[matrix < 5] = 0                                                                                                          In [8]: matrix                                                                                                                        Out[8]: array([[9, 6, 0],        [6, 7, 0],        [7, 0, 0]])

If you want to do this in Python, you can nest loops.

1 2 3 4 5 6 7 8 9 10 11

def set_zero(matrix, tolerance):     result = []     for row in matrix:         cols = []         for value in row:             if value < tolerance:                 cols.append(0)             else:                 cols.append(value)         result.append(cols)     return result

[jefsummers]

1 2 3 4 5 6 7 8 9 10 11

def set_zero(matrix, tolerance):     result = []     for row in matrix:         cols = []         for value in row:             if abs(value) < tolerance:                 cols.append(0)             else:                 cols.append(value)         result.append(cols)     return result

One correction - use abs()