seqbuild 0.0.40__tar.gz → 0.0.41__tar.gz

{seqbuild-0.0.40 → seqbuild-0.0.41}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: seqbuild
-Version: 0.0.40
+Version: 0.0.41
 Summary: A collection of algorithms for pseudo-random number generation.
 Home-page: https://github.com/Changissnz/seqbuild
 Author: Richard Pham

{seqbuild-0.0.40 → seqbuild-0.0.41}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "seqbuild"
-version = "0.0.40"
+version = "0.0.41"
 dependencies = [
 "numpy>=1.22.1",
 "matplotlib>=3.5.3",

{seqbuild-0.0.40 → seqbuild-0.0.41}/seqbuild/desi/differentials.py

@@ -61,7 +61,7 @@ class PIDValueOutputter(GenericIntSeqOp):
 
         if type(self.pic) == type(None) or self.adder_i > self.adder_end:
             self.pts = LPSValueOutputter.new_pointset(self.px,self.l_out)
-            fvec = [self.f_out() % 2 for  _ in range(len(self.pts) - 1)] 
+            fvec = [int(self.f_out()) % 2 for  _ in range(len(self.pts) - 1)] 
             self.pic = PointInterpolationContainer(self.pts,fvec) 
 
             self.adder_end = modulo_in_range(self.f_out(),\

{seqbuild-0.0.40 → seqbuild-0.0.41}/seqbuild/intigers/extraneous.py

@@ -128,9 +128,11 @@ def safe_div(V1,V2):
         y = value_at_index(i,False) 
         q.append(zero_div0(x,y)) 
     return q 
-    
+
+# NOTE: revised to be for uint32 
 def safe_npint32_value(v):
-    r = (np.iinfo(np.int32).min,np.iinfo(np.int32).max)
+    ##r = (np.iinfo(np.int32).min,np.iinfo(np.int32).max)
+    r = (0,np.iinfo(np.int32).max)
     if v >= r[0] and v <= r[1]: return np.int32(v) 
     v_ = modulo_in_range(v,r) 
     return np.int32(v_)

{seqbuild-0.0.40 → seqbuild-0.0.41}/seqbuild/intigers/seq_struct.py

@@ -410,7 +410,7 @@ class ModuloDecompRepr:
         # BUG: strange case: empty, no shift 
         if type(index) == type(None): 
             print("STRANGEE")
-            return qx2 
+            return deepcopy(self.afs_prt) 
 
         # adjust the rest 
         while len(qx) > 0: 

{seqbuild-0.0.40 → seqbuild-0.0.41}/seqbuild/seqgen/ssi_load.py

@@ -355,7 +355,7 @@ class SSIBatchLoader__TypeLCGNet:
             columnOrder=column_order, SSIHop = ssihops,cycleOn = False, cycleIs = 0)
 
     def one_operator_pair(self):
-        base_ops = DEFAULT_PAIRWISE_OPS + [mod]
+        base_ops = DEFAULT_PAIRWISE_OPS + [safe_mod] 
 
         q = prg__one_weighted_pairwise_operator(self.aux_prg,base_ops,\
             base_ops,weight_range=(-2.222,2.555))

{seqbuild-0.0.40 → seqbuild-0.0.41}/seqbuild.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: seqbuild
-Version: 0.0.40
+Version: 0.0.41
 Summary: A collection of algorithms for pseudo-random number generation.
 Home-page: https://github.com/Changissnz/seqbuild
 Author: Richard Pham

{seqbuild-0.0.40 → seqbuild-0.0.41}/setup.py

@@ -5,7 +5,7 @@ with open('README.md') as f:
 
 setup(
     name='seqbuild',
-    version='0.0.40',
+    version='0.0.41',
     description='A collection of algorithms for pseudo-random number generation.',
     long_description=readme,
     author='Richard Pham',

seqbuild-0.0.40/seqbuild/desi/approximators.py

@@ -1,162 +0,0 @@
-from .not_equals import * 
-from .fraction import * 
-from morebs2.poly_interpolation import * 
-
-DEFAULT_FIT22_PARTITION_SIZE_RANGE = (5,103) 
-
-DEFAULT_POLYNOMIAL_PARTITION_SIZE_RANGE = (50,1000)
-DEFAULT_POLYNOMIAL_POINT_SIZE_RANGE = (5,21) 
-
-class Fit22ValueOutputter(GenericIntSeqOp): 
-
-    def __init__(self,px,length_outputter,range_outputter,bool_outputter,\
-            point_conn_type:int,adjustment_type:int=1):  
-        assert type(px) in {MethodType,FunctionType}
-        assert point_conn_type in {1,2}
-        assert type(bool_outputter) in {MethodType,FunctionType}
-
-        self.px = px 
-        self.point_conn_type = point_conn_type
-        self.b_out = bool_outputter
-
-        self.p0 = None
-        self.p1 = None
-        self.H = None
-
-        self.x_ = None
-        self.x2_ = None
-        self.a_ = None
-        self.ax = None 
-        
-        super().__init__(length_outputter,range_outputter,adjustment_type)
-        return
-    
-    def set_next_value(self):
-        if type(self.x_) == type(None):
-            self.update_points() 
-        elif self.x_ > self.x2_: 
-            self.update_points() 
-        else: pass 
-
-        if self.ax: 
-            sx = self.H.f(self.x_)
-        else: 
-            sx = self.H.g(self.x_) 
-        self.x_ += self.a_ 
-
-        sx_ = float_to_string(sx,True,True)
-        self.s = sx_ 
-        return 
-    
-    def update_points(self):
-        # case: start 
-        if type(self.p0) == type(None):
-            self.p0 = self.px() 
-            self.p1 = self.px() 
-        # output 1 or 2 new points 
-        elif self.point_conn_type == 1:
-            self.p0 = self.p1
-            self.p1 = self.px()
-        else:
-            self.p0,self.p1 = self.px(),self.px() 
-
-        self.fix_equal_points()
-
-        # instantiate either <LogFit22> or <Exp2Fit22> 
-        mx = np.array([self.p0,self.p1]) if \
-            self.p0[1] < self.p1[1] else \
-            np.array([self.p1,self.p0])
-        lORe = self.b_out() % 2
-
-        if lORe:
-            self.H = LogFit22(mx,direction=[0,1]) 
-        else: 
-            self.H = Exp2Fit22(mx,direction=[0,1])
-
-        # set delta variables for exploring the fitter 
-        self.ax = self.b_out() % 2 
-
-        ax_ = (self.ax + 1) % 2        
-        self.x_ = min(self.H.ps[:,ax_])
-        self.x2_ = max(self.H.ps[:,ax_]) 
-
-        ap = modulo_in_range(self.l_out(),DEFAULT_FIT22_PARTITION_SIZE_RANGE) 
-        self.a_ = (self.x2_ - self.x_) / ap
-
-    
-    def fix_equal_points(self):
-
-        self.p0 = np.array(self.p0)
-        self.p1 = np.array(self.p1)
-
-        self.p0,self.p1 = \
-            not_equals__pairvec(self.p0,self.p1,self.px,indices=None)
-
-
-
-"""
-LPS stands for Lagrange Polynomial Solver, interpolation of a 
-polynomial's span of points using the Lagrange basis.  
-"""
-class LPSValueOutputter(GenericIntSeqOp):
-
-    def __init__(self,px,length_outputter,range_outputter,\
-        adjustment_type:int=1): 
-        
-        assert type(px) in {MethodType,FunctionType}
-        self.px = px 
-
-        self.p0,self.p1 = None,None 
-        self.adder_end,self.adder,self.adder_i = None,None,None
-
-        super().__init__(length_outputter,range_outputter,adjustment_type)
-
-    #--------------------------------------------------------------------------
-
-    def set_next_value(self):
-        if type(self.p0) == type(None):
-            self.update_points() 
-        elif self.adder_i > self.adder_end: 
-            self.update_points() 
-        else: pass 
-
-        x = self.p0 + (self.adder_i * self.adder)
-        sx = self.lps.output_by_lagrange_basis(x)
-        sx_ = float_to_string(sx,True,True)
-        self.s = sx_ 
-
-        self.adder_i += 1 
-
-        return 
-    
-    def update_points(self):
-
-        self.q = LPSValueOutputter.new_pointset(self.px,self.l_out) 
-
-        # assumes q is a proper set of points for 
-        # fitting via <LagrangePolySolver>
-        self.lps = LagrangePolySolver(self.q, prefetch = True)
-        self.p0,self.p1 = self.lps.minumum,self.lps.maximum 
-
-        self.adder_end = modulo_in_range(self.l_out(),DEFAULT_POLYNOMIAL_PARTITION_SIZE_RANGE)
-        self.adder = (self.p1 - self.p0) / self.adder_end 
-        self.adder_i = 0   
-
-    @staticmethod 
-    def new_pointset(px,l_out):
-
-        num_points = modulo_in_range(\
-            l_out(),DEFAULT_POLYNOMIAL_POINT_SIZE_RANGE) 
-
-        M = []
-        for _ in range(num_points):
-            x = px() 
-            assert type(x) in {list,tuple} or is_vector(x) 
-            assert len(x) == 2 
-            M.append(x) 
-        M = np.array(M) 
-
-        submat_type = l_out() % 4 
-        t = DEFAULT_SUBMAT_TYPES[submat_type]
-        return not_equals__matrix_whole(M,l_out,t)
-