icebug 12.0__tar.gz → 12.2__tar.gz

{icebug-12.0/icebug.egg-info → icebug-12.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: icebug
-Version: 12.0
+Version: 12.2
 Summary: High performance graph analytics backed by read-only memory
 Home-page: https://github.com/Ladybug-Memory/icebug
 Download-URL: https://pypi.python.org/pypi/icebug
@@ -58,6 +58,25 @@ Description-Content-Type: text/markdown
 Requires-Dist: scipy
 Requires-Dist: numpy
 Requires-Dist: pyarrow
+Provides-Extra: test
+Requires-Dist: pytest; extra == "test"
+Requires-Dist: pandas; extra == "test"
+Requires-Dist: networkx; extra == "test"
+Provides-Extra: vizbridges
+Requires-Dist: ipycytoscape; extra == "vizbridges"
+Requires-Dist: plotly; extra == "vizbridges"
+Requires-Dist: seaborn; extra == "vizbridges"
+Requires-Dist: anywidget; extra == "vizbridges"
+Requires-Dist: tabulate; extra == "vizbridges"
+Provides-Extra: pytest
+Requires-Dist: pytest; extra == "pytest"
+Requires-Dist: pandas; extra == "pytest"
+Requires-Dist: networkx; extra == "pytest"
+Requires-Dist: ipycytoscape; extra == "pytest"
+Requires-Dist: plotly; extra == "pytest"
+Requires-Dist: seaborn; extra == "pytest"
+Requires-Dist: anywidget; extra == "pytest"
+Requires-Dist: tabulate; extra == "pytest"
 Dynamic: author-email
 Dynamic: download-url
 Dynamic: home-page
@@ -115,22 +134,9 @@ or build the Python module from source.
 
 While the most recent version is in general available for all package managers, the number of older downloadable versions differ. 
 
-##### pip
+##### uv
 
-    pip3 install [--user] icebug
-
-##### conda (channel conda-forge)
-
-    conda config --add channels conda-forge
-    conda install icebug [-c conda-forge]
-
-##### brew
-
-    brew install icebug
-
-##### spack
-
-    spack install py-icebug
+    uv pip install icebug
 
 More system-specific information on how to install Icebug on Linux, macOS (both Intel and M1) and Windows-systems can be found [here](https://networkit.github.io/get_started.html).
 
@@ -155,7 +161,7 @@ We also provide a Binder-instance of our notebooks. To access this service, you
 
 If you only want to see in short how Icebug is used - the following example provides a climpse at that. Here we generate a random hyperbolic graph with 100k nodes and compute its communities with the PLM method:
 
-    >>> import icebug as nk
+    >>> import networkit as nk
     >>> g = nk.generators.HyperbolicGenerator(1e5).generate()
     >>> communities = nk.community.detectCommunities(g, inspect=True)
     PLM(balanced,pc,turbo) detected communities in 0.14577102661132812 [s]

{icebug-12.0 → icebug-12.2}/README.md

@@ -50,22 +50,9 @@ or build the Python module from source.
 
 While the most recent version is in general available for all package managers, the number of older downloadable versions differ. 
 
-##### pip
+##### uv
 
-    pip3 install [--user] icebug
-
-##### conda (channel conda-forge)
-
-    conda config --add channels conda-forge
-    conda install icebug [-c conda-forge]
-
-##### brew
-
-    brew install icebug
-
-##### spack
-
-    spack install py-icebug
+    uv pip install icebug
 
 More system-specific information on how to install Icebug on Linux, macOS (both Intel and M1) and Windows-systems can be found [here](https://networkit.github.io/get_started.html).
 
@@ -90,7 +77,7 @@ We also provide a Binder-instance of our notebooks. To access this service, you
 
 If you only want to see in short how Icebug is used - the following example provides a climpse at that. Here we generate a random hyperbolic graph with 100k nodes and compute its communities with the PLM method:
 
-    >>> import icebug as nk
+    >>> import networkit as nk
     >>> g = nk.generators.HyperbolicGenerator(1e5).generate()
     >>> communities = nk.community.detectCommunities(g, inspect=True)
     PLM(balanced,pc,turbo) detected communities in 0.14577102661132812 [s]

{icebug-12.0 → icebug-12.2/icebug.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: icebug
-Version: 12.0
+Version: 12.2
 Summary: High performance graph analytics backed by read-only memory
 Home-page: https://github.com/Ladybug-Memory/icebug
 Download-URL: https://pypi.python.org/pypi/icebug
@@ -58,6 +58,25 @@ Description-Content-Type: text/markdown
 Requires-Dist: scipy
 Requires-Dist: numpy
 Requires-Dist: pyarrow
+Provides-Extra: test
+Requires-Dist: pytest; extra == "test"
+Requires-Dist: pandas; extra == "test"
+Requires-Dist: networkx; extra == "test"
+Provides-Extra: vizbridges
+Requires-Dist: ipycytoscape; extra == "vizbridges"
+Requires-Dist: plotly; extra == "vizbridges"
+Requires-Dist: seaborn; extra == "vizbridges"
+Requires-Dist: anywidget; extra == "vizbridges"
+Requires-Dist: tabulate; extra == "vizbridges"
+Provides-Extra: pytest
+Requires-Dist: pytest; extra == "pytest"
+Requires-Dist: pandas; extra == "pytest"
+Requires-Dist: networkx; extra == "pytest"
+Requires-Dist: ipycytoscape; extra == "pytest"
+Requires-Dist: plotly; extra == "pytest"
+Requires-Dist: seaborn; extra == "pytest"
+Requires-Dist: anywidget; extra == "pytest"
+Requires-Dist: tabulate; extra == "pytest"
 Dynamic: author-email
 Dynamic: download-url
 Dynamic: home-page
@@ -115,22 +134,9 @@ or build the Python module from source.
 
 While the most recent version is in general available for all package managers, the number of older downloadable versions differ. 
 
-##### pip
+##### uv
 
-    pip3 install [--user] icebug
-
-##### conda (channel conda-forge)
-
-    conda config --add channels conda-forge
-    conda install icebug [-c conda-forge]
-
-##### brew
-
-    brew install icebug
-
-##### spack
-
-    spack install py-icebug
+    uv pip install icebug
 
 More system-specific information on how to install Icebug on Linux, macOS (both Intel and M1) and Windows-systems can be found [here](https://networkit.github.io/get_started.html).
 
@@ -155,7 +161,7 @@ We also provide a Binder-instance of our notebooks. To access this service, you
 
 If you only want to see in short how Icebug is used - the following example provides a climpse at that. Here we generate a random hyperbolic graph with 100k nodes and compute its communities with the PLM method:
 
-    >>> import icebug as nk
+    >>> import networkit as nk
     >>> g = nk.generators.HyperbolicGenerator(1e5).generate()
     >>> communities = nk.community.detectCommunities(g, inspect=True)
     PLM(balanced,pc,turbo) detected communities in 0.14577102661132812 [s]

icebug-12.2/icebug.egg-info/requires.txt

@@ -0,0 +1,25 @@
+scipy
+numpy
+pyarrow
+
+[pytest]
+pytest
+pandas
+networkx
+ipycytoscape
+plotly
+seaborn
+anywidget
+tabulate
+
+[test]
+pytest
+pandas
+networkx
+
+[vizbridges]
+ipycytoscape
+plotly
+seaborn
+anywidget
+tabulate

{icebug-12.0 → icebug-12.2}/include/networkit/graph/Graph.hpp

@@ -157,6 +157,10 @@ protected:
     //!< Arrow array for CSR indptr (offsets into indices array) for incoming edges - only for
     //!< directed graphs
     std::shared_ptr<arrow::UInt64Array> inEdgesCSRIndptr;
+    //!< Arrow array for CSR edge weights for outgoing edges
+    std::shared_ptr<arrow::DoubleArray> outEdgesCSRWeights;
+    //!< Arrow array for CSR edge weights for incoming edges - only for directed graphs
+    std::shared_ptr<arrow::DoubleArray> inEdgesCSRWeights;
     //!< flag to indicate if CSR arrays are being used instead of vectors
     bool usingCSR;
 
@@ -754,7 +758,8 @@ public:
           directed(other.directed), edgesIndexed(other.edgesIndexed), deletedID(other.deletedID),
           exists(other.exists), outEdgesCSRIndices(other.outEdgesCSRIndices),
           outEdgesCSRIndptr(other.outEdgesCSRIndptr), inEdgesCSRIndices(other.inEdgesCSRIndices),
-          inEdgesCSRIndptr(other.inEdgesCSRIndptr), usingCSR(other.usingCSR),
+          inEdgesCSRIndptr(other.inEdgesCSRIndptr), outEdgesCSRWeights(other.outEdgesCSRWeights),
+          inEdgesCSRWeights(other.inEdgesCSRWeights), usingCSR(other.usingCSR),
           // call special constructors to copy attribute maps
           nodeAttributeMap(other.nodeAttributeMap, this),
           edgeAttributeMap(other.edgeAttributeMap, this) {
@@ -779,7 +784,8 @@ protected:
           directed(other.directed), edgesIndexed(other.edgesIndexed), deletedID(other.deletedID),
           exists(other.exists), outEdgesCSRIndices(other.outEdgesCSRIndices),
           outEdgesCSRIndptr(other.outEdgesCSRIndptr), inEdgesCSRIndices(other.inEdgesCSRIndices),
-          inEdgesCSRIndptr(other.inEdgesCSRIndptr), usingCSR(other.usingCSR),
+          inEdgesCSRIndptr(other.inEdgesCSRIndptr), outEdgesCSRWeights(other.outEdgesCSRWeights),
+          inEdgesCSRWeights(other.inEdgesCSRWeights), usingCSR(other.usingCSR),
           // call special constructors to copy attribute maps
           nodeAttributeMap(other.nodeAttributeMap, this),
           edgeAttributeMap(other.edgeAttributeMap, this) {
@@ -796,7 +802,9 @@ public:
           exists(std::move(other.exists)), outEdgesCSRIndices(std::move(other.outEdgesCSRIndices)),
           outEdgesCSRIndptr(std::move(other.outEdgesCSRIndptr)),
           inEdgesCSRIndices(std::move(other.inEdgesCSRIndices)),
-          inEdgesCSRIndptr(std::move(other.inEdgesCSRIndptr)), usingCSR(other.usingCSR),
+          inEdgesCSRIndptr(std::move(other.inEdgesCSRIndptr)),
+          outEdgesCSRWeights(std::move(other.outEdgesCSRWeights)),
+          inEdgesCSRWeights(std::move(other.inEdgesCSRWeights)), usingCSR(other.usingCSR),
           nodeAttributeMap(std::move(other.nodeAttributeMap)),
           edgeAttributeMap(std::move(other.edgeAttributeMap)) {
         // attributes: set graph pointer to this new graph
@@ -816,11 +824,16 @@ public:
      * directed graphs).
      * @param inIndptr Arrow array containing offsets into inIndices for each node (only for
      * directed graphs).
+     * @param outWeights Arrow array containing edge weights for outgoing edges (optional).
+     * @param inWeights Arrow array containing edge weights for incoming edges (optional, only for
+     * directed graphs).
      */
     Graph(count n, bool directed, std::shared_ptr<arrow::UInt64Array> outIndices,
           std::shared_ptr<arrow::UInt64Array> outIndptr,
           std::shared_ptr<arrow::UInt64Array> inIndices = nullptr,
-          std::shared_ptr<arrow::UInt64Array> inIndptr = nullptr);
+          std::shared_ptr<arrow::UInt64Array> inIndptr = nullptr,
+          std::shared_ptr<arrow::DoubleArray> outWeights = nullptr,
+          std::shared_ptr<arrow::DoubleArray> inWeights = nullptr);
 
     /** move assignment operator */
     Graph &operator=(Graph &&other) noexcept {
@@ -839,6 +852,8 @@ public:
         std::swap(outEdgesCSRIndptr, other.outEdgesCSRIndptr);
         std::swap(inEdgesCSRIndices, other.inEdgesCSRIndices);
         std::swap(inEdgesCSRIndptr, other.inEdgesCSRIndptr);
+        std::swap(outEdgesCSRWeights, other.outEdgesCSRWeights);
+        std::swap(inEdgesCSRWeights, other.inEdgesCSRWeights);
         std::swap(deletedID, other.deletedID);
 
         // attributes: set graph pointer to this new graph
@@ -1710,12 +1725,11 @@ inline void Graph::forInEdgesOfImpl(node u, L handle) const {
                 // Get edge ID (CSR graphs don't currently support edge IDs)
                 edgeid eid = graphHasEdgeIds ? none : none;
 
-                // For incoming edges, v is the source and u is the target
-                edgeLambda(handle, v, u, weight, eid);
+                // For incoming edges: u is current node (target), v is neighbor (source)
+                edgeLambda(handle, u, v, weight, eid);
             }
         } else {
             // For undirected graphs, incoming edges are the same as outgoing edges
-            // but we need to swap u and v in the handle call
             auto [neighbors, degree] = getCSROutNeighbors(u);
             if (neighbors == nullptr || degree == 0)
                 return;
@@ -1730,8 +1744,8 @@ inline void Graph::forInEdgesOfImpl(node u, L handle) const {
                 // Get edge ID (CSR graphs don't currently support edge IDs)
                 edgeid eid = graphHasEdgeIds ? none : none;
 
-                // For undirected graphs, call with v, u (swapped)
-                edgeLambda(handle, v, u, weight, eid);
+                // For undirected graphs, pass (u, v) where u is current node and v is neighbor
+                edgeLambda(handle, u, v, weight, eid);
             }
         }
     } else {
@@ -1739,16 +1753,12 @@ inline void Graph::forInEdgesOfImpl(node u, L handle) const {
         // Check exists for mutable graphs
         if (!exists[u])
             return;
-        // Create a wrapper function that swaps the first two arguments
-        // The handle expects (source, target, ...) but virtual method provides (target, source,
-        // ...)
-        auto wrapper = std::function<void(node, node, edgeweight, edgeid)>(
-            [&](node target, node source, edgeweight w, edgeid e) {
-                // Swap arguments: call handle with (source, target, ...) instead of (target,
-                // source, ...)
-                edgeLambda(handle, source, target, w, e);
-            });
-        forInEdgesVirtualImpl(u, graphIsDirected, hasWeights, graphHasEdgeIds, wrapper);
+        // GraphW's forInEdgesVirtualImpl calls handle(v, u, ...) where v is neighbor and u is
+        // current node We need to swap so that edgeLambda receives (current, neighbor, ...) and
+        // passes neighbor to f
+        forInEdgesVirtualImpl(
+            u, graphIsDirected, hasWeights, graphHasEdgeIds,
+            [&](node v, node u, edgeweight w, edgeid e) { edgeLambda(handle, u, v, w, e); });
     }
 }
 

{icebug-12.0 → icebug-12.2}/include/networkit/graph/GraphR.hpp

@@ -51,13 +51,18 @@ public:
      * directed graphs).
      * @param inIndptr Arrow array containing offsets into inIndices for each node (only for
      * directed graphs).
+     * @param outWeights Arrow array containing edge weights for outgoing edges (optional).
+     * @param inWeights Arrow array containing edge weights for incoming edges (optional, only for
+     * directed graphs).
      */
     GraphR(count n, bool directed, std::shared_ptr<arrow::UInt64Array> outIndices,
            std::shared_ptr<arrow::UInt64Array> outIndptr,
            std::shared_ptr<arrow::UInt64Array> inIndices = nullptr,
-           std::shared_ptr<arrow::UInt64Array> inIndptr = nullptr)
+           std::shared_ptr<arrow::UInt64Array> inIndptr = nullptr,
+           std::shared_ptr<arrow::DoubleArray> outWeights = nullptr,
+           std::shared_ptr<arrow::DoubleArray> inWeights = nullptr)
         : Graph(n, directed, std::move(outIndices), std::move(outIndptr), std::move(inIndices),
-                std::move(inIndptr)) {}
+                std::move(inIndptr), std::move(outWeights), std::move(inWeights)) {}
 
     /**
      * Copy constructor
@@ -96,7 +101,8 @@ public:
 
     /**
      * Return edge weight of edge {@a u,@a v}. Returns 0 if edge does not
-     * exist. For CSR graphs, always returns 1.0 if edge exists (unweighted).
+     * exist. If weights are provided during construction, returns the actual
+     * edge weight; otherwise returns defaultEdgeWeight (1.0).
      *
      * @param u Endpoint of edge.
      * @param v Endpoint of edge.

{icebug-12.0 → icebug-12.2}/include/networkit/graph/GraphW.hpp

@@ -96,6 +96,8 @@ public:
         outEdgesCSRIndptr.reset();
         inEdgesCSRIndices.reset();
         inEdgesCSRIndptr.reset();
+        outEdgesCSRWeights.reset();
+        inEdgesCSRWeights.reset();
         usingCSR = false;
     }
 
@@ -116,7 +118,14 @@ public:
     template <class EdgeMerger = std::plus<edgeweight>>
     GraphW(const Graph &other, bool weighted, bool directed, bool edgesIndexed = false,
            EdgeMerger edgeMerger = std::plus<edgeweight>())
-        : GraphW(other.numberOfNodes(), weighted, directed, edgesIndexed) {
+        : GraphW(other.upperNodeIdBound(), weighted, directed, edgesIndexed) {
+
+        // Remove nodes that don't exist in the source graph
+        for (node v = 0; v < other.upperNodeIdBound(); ++v) {
+            if (!other.hasNode(v)) {
+                removeNode(v);
+            }
+        }
 
         // Copy all edges using the public API
         if (other.isDirected() == directed) {

{icebug-12.0 → icebug-12.2}/networkit/__init__.py

@@ -30,7 +30,7 @@ __credits__ = ["Eugenio Angriman", "Lukas Barth", "Miriam Beddig", "Elisabetta B
     "Moritz v. Looz", "Yassine Marrakchi", "Henning Meyerhenke", "Manuel Penschuck", "Lucas Archimedes Gregorio Henr Petersen", "Marcel Radermacher", "Klara Reichard", \
 	"Marvin Ritter", "Aleksejs Sazonovs", "Hung Tran", "Alexander van der Grinten", "Florian Weber", "Michael Wegner", "Jörg Weisbarth"]
 __license__ = "MIT"
-__version__ = "12.0"
+__version__ = "12.2"
 
 # standard library modules
 import csv

{icebug-12.0 → icebug-12.2}/networkit/algebraic.py

@@ -7,11 +7,12 @@ __author__ = "Christian Staudt"
 
 # external imports
 import scipy.sparse
+import scipy.linalg
 from scipy.sparse import csgraph, linalg
 import numpy as np
 
 def column(matrix, i):
-	""" 
+	"""
 	column(matrix, i)
 
 	Get the ith column of a matrix
@@ -27,12 +28,12 @@ def column(matrix, i):
 	-------
 	list(float)
 		Column i of matrix.
-	"""	
+	"""
 	return [row[i] for row in matrix]
 
 
 def adjacencyMatrix(G, matrixType="sparse"):
-	""" 
+	"""
 	adjacencyMatrix(G, matrixType="sparse")
 
 	Get the adjacency matrix of the graph `G`.
@@ -79,7 +80,7 @@ def adjacencyMatrix(G, matrixType="sparse"):
 	return A
 
 def laplacianMatrix(G):
-	""" 
+	"""
 	laplacianMatrix(G)
 
 	Get the laplacian matrix of the graph `G`.
@@ -200,15 +201,38 @@ def eigenvectors(matrix, cutoff=-1, reverse=False):
 	if cutoff == -1:
 		cutoff = matrix.shape[0] - 3
 
-	if reverse:
-		mode = "SR"
+	# For small matrices or when requesting many eigenvalues, use dense solver
+	# as sparse solver can be unreliable
+	if matrix.shape[0] <= 10 or cutoff + 1 >= matrix.shape[0] - 1:
+		# Convert to dense and use scipy.linalg.eig
+		w, v = scipy.linalg.eig(matrix.toarray())
+
+		# Convert to numpy complex128 for consistent sorting behavior
+		w = np.array(w, dtype=np.complex128)
+
+		# Sort all eigenvalues by real part (ascending)
+		orderlist = list(zip(w, range(0, len(w))))
+		orderlist = sorted(orderlist)
+
+		# Select the requested eigenvalues based on mode
+		if reverse:
+			# mode='SR': want smallest real part
+			orderlist = orderlist[:cutoff + 1]
+		else:
+			# mode='LR': want largest real part
+			orderlist = orderlist[-(cutoff + 1):]
+			# Re-sort to maintain ascending order
+			orderlist = sorted(orderlist)
 	else:
-		mode = "LR"
+		if reverse:
+			mode = "SR"
+		else:
+			mode = "LR"
 
-	w, v = scipy.sparse.linalg.eigs(matrix, cutoff + 1, which=mode)
+		w, v = scipy.sparse.linalg.eigs(matrix, cutoff + 1, which=mode)
 
-	orderlist = zip(w, range(0, len(w)))
-	orderlist = sorted(orderlist)
+		orderlist = zip(w, range(0, len(w)))
+		orderlist = sorted(orderlist)
 
 	orderedW = column(orderlist, 0)
 	orderedV = [v[:,i] for i in column(orderlist, 1)]