Merge pull request #169 from kandeeban15/master

Update Python Version to 3.11.3

CMakeLists.txt

@@ -41,7 +41,7 @@ add_executable(JasmineGraph main.cpp src/server/JasmineGraphServer.cpp src/serve
         src/frontend/core/factory/ExecutorFactory.cpp src/frontend/core/factory/ExecutorFactory.h src/frontend/core/CoreConstants.cpp
         src/frontend/core/CoreConstants.h)
 
-target_compile_options(JasmineGraph PRIVATE -std=c++11 -I/usr/include/python3.5)
+target_compile_options(JasmineGraph PRIVATE -std=c++11 -I/usr/include/python3.11)
 
 if(CMAKE_ENABLE_DEBUG)
     message(STATUS "DEBUG enabled")
@@ -58,11 +58,11 @@ target_link_libraries(JasmineGraph /usr/lib/x86_64-linux-gnu/libflatbuffers.a)
 target_link_libraries(JasmineGraph /usr/lib/x86_64-linux-gnu/libjsoncpp.a)
 target_link_libraries(JasmineGraph /usr/local/lib/libcppkafka.so)
 
-set(PYTHON_EXECUTABLE "/usr/bin/python3.5")
-set(PYTHON_INCLUDE_DIR "/usr/include/python3.5m")
-set(PYTHON_LIBRARIES "/usr/lib/x86_64-linux-gnu/libpython3.5m.so")
+set(PYTHON_EXECUTABLE "/usr/bin/python3.11")
+set(PYTHON_INCLUDE_DIR "/usr/include/python3.11m")
+set(PYTHON_LIBRARIES "/usr/lib/x86_64-linux-gnu/libpython3.11.so")
 
 include_directories(${PYTHON_INCLUDE_DIRS})
 include_directories(${PYTHON_DIRECTORIES})
 target_link_libraries(JasmineGraph ${PYTHON_LIBRARIES})
-target_link_libraries(JasmineGraph m)
+target_link_libraries(JasmineGraph m)

Dockerfile

@@ -7,10 +7,13 @@ RUN apt-get update
 RUN apt-get install --no-install-recommends -y apt-transport-https
 RUN apt-get update
 RUN apt-get install --no-install-recommends -y curl gnupg2 ca-certificates software-properties-common nlohmann-json3-dev
-RUN apt-get install --no-install-recommends -y git cmake build-essential sqlite3 libsqlite3-dev libssl-dev librdkafka-dev libboost-all-dev libtool libxerces-c-dev libflatbuffers-dev python3-pip
+
+RUN apt-get install --no-install-recommends -y git cmake build-essential sqlite3 libsqlite3-dev libssl-dev librdkafka-dev libboost-all-dev libtool libxerces-c-dev libflatbuffers-dev
 RUN add-apt-repository ppa:deadsnakes/ppa
-RUN apt-get install --no-install-recommends -y python3.5-dev
+RUN apt-get install --no-install-recommends -y python3.11-dev
+RUN curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py
 RUN apt-get install --no-install-recommends -y libjsoncpp-dev libspdlog-dev pigz
+RUN python3.11 get-pip.py
 
 RUN curl -fsSL https://download.docker.com/linux/debian/gpg | apt-key add -
 RUN add-apt-repository "deb [arch=amd64] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable"
@@ -38,10 +41,10 @@ ENV JASMINEGRAPH_HOME="/home/ubuntu/software/jasminegraph"
 RUN mkdir /home/ubuntu/software/jasminegraph
 WORKDIR /home/ubuntu/software/jasminegraph
 
-RUN pip install tensorflow==2.5.3
-RUN pip install -U scikit-learn
 COPY ./GraphSAGE ./GraphSAGE
 RUN pip install -r ./GraphSAGE/requirements
+RUN pip install joblib
+RUN pip install threadpoolctl
 
 COPY ./conf ./conf
 COPY ./src ./src
@@ -54,4 +57,4 @@ COPY ./run-docker.sh ./run-docker.sh
 
 RUN sh build.sh
 ENTRYPOINT ["/home/ubuntu/software/jasminegraph/run-docker.sh"]
-CMD ["bash"]
+CMD ["bash"]

GraphSAGE/graphsage/aggregators.py

@@ -1,8 +1,10 @@
-import tensorflow as tf
+import tensorflow.compat.v1 as tf
 
 from layers import Layer, Dense
 from inits import glorot, zeros
 
+tf.disable_v2_behavior()
+
 class MeanAggregator(Layer):
     """
     Aggregates via mean followed by matmul and non-linearity.

GraphSAGE/graphsage/inits.py

@@ -1,6 +1,8 @@
-import tensorflow as tf
+import tensorflow.compat.v1 as tf
 import numpy as np
 
+tf.disable_v2_behavior()
+
 # DISCLAIMER:
 # Parts of this code file are derived from
 # https://github.com/tkipf/gcn

GraphSAGE/graphsage/layers.py

@@ -1,10 +1,11 @@
 from __future__ import division
 from __future__ import print_function
 
-import tensorflow as tf
+import tensorflow.compat.v1 as tf
 
 from inits import zeros
 
+tf.disable_v2_behavior()
 flags = tf.app.flags
 FLAGS = flags.FLAGS
 

GraphSAGE/graphsage/metrics.py

@@ -1,4 +1,5 @@
-import tensorflow as tf
+import tensorflow.compat.v1 as tf
+tf.disable_v2_behavior()
 
 # DISCLAIMER:
 # Parts of this code file were originally forked from

GraphSAGE/graphsage/minibatch.py

@@ -25,16 +25,16 @@ def __init__(self, G, G_local,id2idx,
             **kwargs):
 
         self.G = G
-        self.nodes = G.nodes()
+        self.nodes = list(G.nodes())
         self.local_nodes = G_local.nodes()
-        self.central_nodes = list(set(G.nodes())-set(G_local.nodes()))
+        self.central_nodes = list(set(list(G.nodes()))-set(G_local.nodes()))
         self.id2idx = id2idx
         self.placeholders = placeholders
         self.batch_size = batch_size
         self.max_degree = max_degree
         self.batch_num = 0
 
-        self.nodes = np.random.permutation(G.nodes())
+        self.nodes = np.random.permutation(list(G.nodes()))
         self.adj, self.deg = self.construct_adj()
         self.test_adj = self.construct_test_adj()
         if context_pairs is None:
@@ -52,12 +52,12 @@ def __init__(self, G, G_local,id2idx,
             else:
                 self.train_edges = self.val_edges = self.edges
 
-        print(len([n for n in G.nodes() if not G.node[n]['test'] and not G.node[n]['val']]), 'train nodes')
-        print(len([n for n in G.nodes() if G.node[n]['test'] or G.node[n]['val']]), 'test nodes')
+        print(len([n for n in list(G.nodes()) if not G.nodes[n]['test'] and not G.nodes[n]['val']]), 'train nodes')
+        print(len([n for n in list(G.nodes()) if G.nodes[n]['test'] or G.nodes[n]['val']]), 'test nodes')
         self.val_set_size = len(self.val_edges)
 
     def _n2v_prune(self, edges):
-        is_val = lambda n : self.G.node[n]["val"] or self.G.node[n]["test"]
+        is_val = lambda n : self.G.nodes[n]["val"] or self.G.nodes[n]["test"]
         return [e for e in edges if not is_val(e[1])]
 
     def _remove_isolated(self, edge_list):
@@ -67,11 +67,11 @@ def _remove_isolated(self, edge_list):
         count = 0
         for n1, n2 in edge_list:
             count +=1
-            if not n1 in self.G.node or not n2 in self.G.node:
+            if not n1 in list(self.G.nodes()) or not n2 in list(self.G.nodes()):
                 missing += 1
                 continue
-            if (self.G.node[n1]['test'] and self.G.node[n2]['test']) \
-                    or (self.G.node[n1]['val'] and self.G.node[n2]['val']):
+            if (self.G.nodes[n1]['test'] and self.G.nodes[n2]['test']) \
+                    or (self.G.nodes[n1]['val'] and self.G.nodes[n2]['val']):
                 continue
             else:
                 new_edge_list.append((n1,n2))
@@ -82,8 +82,8 @@ def construct_adj(self):
         adj = len(self.id2idx)*np.ones((len(self.id2idx)+1, self.max_degree))
         deg = np.zeros((len(self.id2idx),))
 
-        for nodeid in self.G.nodes():
-            if self.G.node[nodeid]['test'] or self.G.node[nodeid]['val']:
+        for nodeid in list(self.G.nodes()):
+            if self.G.nodes[nodeid]['test'] or self.G.nodes[nodeid]['val']:
                 continue
             neighbors = np.array([self.id2idx[neighbor] 
                 for neighbor in self.G.neighbors(nodeid)
@@ -100,7 +100,7 @@ def construct_adj(self):
 
     def construct_test_adj(self):
         adj = len(self.id2idx)*np.ones((len(self.id2idx)+1, self.max_degree))
-        for nodeid in self.G.nodes():
+        for nodeid in list(self.G.nodes()):
             neighbors = np.array([self.id2idx[neighbor] 
                 for neighbor in self.G.neighbors(nodeid)])
             if len(neighbors) == 0:
@@ -183,8 +183,8 @@ def label_val(self):
         train_edges = []
         val_edges = []
         for n1, n2 in self.G.edges():
-            if (self.G.node[n1]['val'] or self.G.node[n1]['test'] 
-                    or self.G.node[n2]['val'] or self.G.node[n2]['test']):
+            if (self.G.nodes[n1]['val'] or self.G.nodes[n1]['test'] 
+                    or self.G.nodes[n2]['val'] or self.G.nodes[n2]['test']):
                 val_edges.append((n1,n2))
             else:
                 train_edges.append((n1,n2))
@@ -217,7 +217,7 @@ def __init__(self, G, id2idx,
             **kwargs):
 
         self.G = G
-        self.nodes = G.nodes()
+        self.nodes = list(G.nodes())
         self.id2idx = id2idx
         self.placeholders = placeholders
         self.batch_size = batch_size
@@ -229,11 +229,11 @@ def __init__(self, G, id2idx,
         self.adj, self.deg = self.construct_adj()
         self.test_adj = self.construct_test_adj()
 
-        self.val_nodes = [n for n in self.G.nodes() if self.G.node[n]['val']]
-        self.test_nodes = [n for n in self.G.nodes() if self.G.node[n]['test']]
+        self.val_nodes = [n for n in list(self.G.nodes()) if self.G.nodes[n]['val']]
+        self.test_nodes = [n for n in list(self.G.nodes()) if self.G.nodes[n]['test']]
 
         self.no_train_nodes_set = set(self.val_nodes + self.test_nodes)
-        self.train_nodes = set(G.nodes()).difference(self.no_train_nodes_set)
+        self.train_nodes = set(list(G.nodes())).difference(self.no_train_nodes_set)
         # don't train on nodes that only have edges to test set
         self.train_nodes = [n for n in self.train_nodes if self.deg[id2idx[n]] > 0]
 
@@ -251,8 +251,8 @@ def construct_adj(self):
         adj = len(self.id2idx)*np.ones((len(self.id2idx)+1, self.max_degree))
         deg = np.zeros((len(self.id2idx),))
 
-        for nodeid in self.G.nodes():
-            if self.G.node[nodeid]['test'] or self.G.node[nodeid]['val']:
+        for nodeid in list(self.G.nodes()):
+            if self.G.nodes[nodeid]['test'] or self.G.nodes[nodeid]['val']:
                 continue
             neighbors = np.array([self.id2idx[neighbor]
                 for neighbor in self.G.neighbors(nodeid)
@@ -269,7 +269,7 @@ def construct_adj(self):
 
     def construct_test_adj(self):
         adj = len(self.id2idx)*np.ones((len(self.id2idx)+1, self.max_degree))
-        for nodeid in self.G.nodes():
+        for nodeid in list(self.G.nodes()):
             neighbors = np.array([self.id2idx[neighbor] 
                 for neighbor in self.G.neighbors(nodeid)])
             if len(neighbors) == 0:

GraphSAGE/graphsage/models.py

@@ -1,6 +1,6 @@
 from collections import namedtuple
 
-import tensorflow as tf
+import tensorflow.compat.v1 as tf
 import math
 
 import layers as layers
@@ -9,6 +9,7 @@
 from prediction import BipartiteEdgePredLayer
 from aggregators import MeanAggregator, MaxPoolingAggregator, MeanPoolingAggregator, SeqAggregator, GCNAggregator
 
+tf.disable_v2_behavior()
 flags = tf.app.flags
 FLAGS = flags.FLAGS
 

GraphSAGE/graphsage/neigh_samplers.py

@@ -3,7 +3,8 @@
 
 from layers import Layer
 
-import tensorflow as tf
+import tensorflow.compat.v1 as tf
+tf.disable_v2_behavior()
 flags = tf.app.flags
 FLAGS = flags.FLAGS
 

GraphSAGE/graphsage/prediction.py

@@ -3,8 +3,9 @@
 
 from inits import zeros
 from layers import Layer
-import tensorflow as tf
+import tensorflow.compat.v1 as tf
 
+tf.disable_v2_behavior()
 flags = tf.app.flags
 FLAGS = flags.FLAGS
 

GraphSAGE/graphsage/supervised_models.py

@@ -1,9 +1,10 @@
-import tensorflow as tf
+import tensorflow.compat.v1 as tf
 
 import models as models
 import layers as layers
 from aggregators import MeanAggregator, MaxPoolingAggregator, MeanPoolingAggregator, SeqAggregator, GCNAggregator
 
+tf.disable_v2_behavior()
 flags = tf.app.flags
 FLAGS = flags.FLAGS
 

GraphSAGE/graphsage/supervised_train.py

@@ -3,7 +3,7 @@
 
 import os
 import time
-import tensorflow as tf
+import tensorflow.compat.v1 as tf
 import numpy as np
 import sklearn
 from sklearn import metrics
@@ -16,6 +16,7 @@
 
 os.environ["CUDA_DEVICE_ORDER"]="PCI_BUS_ID"
 
+tf.disable_v2_behavior()
 # Set random seed
 seed = 123
 np.random.seed(seed)
@@ -25,7 +26,7 @@
 flags = tf.app.flags
 FLAGS = flags.FLAGS
 
-tf.app.flags.DEFINE_boolean('log_device_placement', False,
+flags.DEFINE_boolean('log_device_placement', False,
                             """Whether to log device placement.""")
 #core params..
 flags.DEFINE_string('model', 'graphsage_mean', 'model names. See README for possible values.')

GraphSAGE/graphsage/unsupervised_train.py

@@ -3,7 +3,7 @@
 
 import os
 import time
-import tensorflow as tf
+import tensorflow.compat.v1 as tf
 import numpy as np
 import datetime
 
@@ -14,16 +14,18 @@
 
 os.environ["CUDA_DEVICE_ORDER"]="PCI_BUS_ID"
 
+tf.disable_v2_behavior()
+
 # Set random seed
 seed = 123
 np.random.seed(seed)
-tf.set_random_seed(seed)
+tf.random.set_random_seed(seed)
 
 # Settings
 flags = tf.app.flags
 FLAGS = flags.FLAGS
 
-tf.app.flags.DEFINE_boolean('log_device_placement', False,
+flags.DEFINE_boolean('log_device_placement', False,
                             """Whether to log device placement.""")
 #core params..
 flags.DEFINE_string('model', 'graphsage_mean', 'model names. See README for possible values.')
@@ -66,7 +68,7 @@
 
 def log_dir():
 
-    log_dir = FLAGS.base_log_dir + "/jasminegraph-local_trained_model_store/"
+    log_dir = FLAGS.base_log_dir + "jasminegraph-local_trained_model_store/"
     log_dir += "{graph_id:s}_model_{train_worker:s}".format(
         graph_id = FLAGS.train_prefix.split("/")[-1],
         train_worker = FLAGS.train_worker
@@ -414,9 +416,9 @@ def train(train_data, G_local,test_data=None):
 
         if FLAGS.model == "n2v":
             # stopping the gradient for the already trained nodes
-            train_ids = tf.constant([[id_map[n]] for n in G.nodes_iter() if not G.node[n]['val'] and not G.node[n]['test']],
+            train_ids = tf.constant([[id_map[n]] for n in G.nodes_iter() if not G._node[n]['val'] and not G._node[n]['test']],
                     dtype=tf.int32)
-            test_ids = tf.constant([[id_map[n]] for n in G.nodes_iter() if G.node[n]['val'] or G.node[n]['test']], 
+            test_ids = tf.constant([[id_map[n]] for n in G.nodes_iter() if G._node[n]['val'] or G._node[n]['test']],
                     dtype=tf.int32)
             update_nodes = tf.nn.embedding_lookup(model.context_embeds, tf.squeeze(test_ids))
             no_update_nodes = tf.nn.embedding_lookup(model.context_embeds,tf.squeeze(train_ids))
@@ -427,7 +429,7 @@ def train(train_data, G_local,test_data=None):
 
             # run random walks
             from graphsage.utils import run_random_walks
-            nodes = [n for n in G.nodes_iter() if G.node[n]["val"] or G.node[n]["test"]]
+            nodes = [n for n in G.nodes_iter() if G._node[n]["val"] or G._node[n]["test"]]
             start_time = time.time()
             pairs = run_random_walks(G, nodes, num_walks=50)
             walk_time = time.time() - start_time