diff options
Diffstat (limited to 'src')
| -rw-r--r-- | src/algorithms.py | 6 | ||||
| -rw-r--r-- | src/cascade_creation.py | 6 | ||||
| -rw-r--r-- | src/convex_optimization.py | 4 | ||||
| -rw-r--r-- | src/make_plots.py | 18 |
4 files changed, 18 insertions, 16 deletions
diff --git a/src/algorithms.py b/src/algorithms.py index 24a63f3..43c6721 100644 --- a/src/algorithms.py +++ b/src/algorithms.py @@ -38,7 +38,7 @@ def recovery_passed_function(G, cascades, floor_cstt, passed_function, passed_function should have similar structure to ones in convex_optimation """ G_hat = cascade_creation.InfluenceGraph(max_proba=None) - G_hat.add_nodes_from(G.nodes()) + G_hat.add_nodes_from(G.nodes()) f_x, f_xz = passed_function(*args, **kwargs) @@ -83,8 +83,8 @@ def correctness_measure(G, G_hat, print_values=True): print("Precision: {}".format(precision)) print("Recall: {}".format(recall)) print("F1 score: {}".format(f1_score)) - print("Fall Out: {}".format(fall_out)) - print("l2_norm: {}".format(norm)) + print("Fall Out: {}".format(fall_out)) + print("l2_norm: {}".format(norm)) return fp, fn, tp, tn diff --git a/src/cascade_creation.py b/src/cascade_creation.py index cdf7484..714c120 100644 --- a/src/cascade_creation.py +++ b/src/cascade_creation.py @@ -65,7 +65,8 @@ class InfluenceGraph(nx.DiGraph): if not hasattr(self, '_mat'): if self.min_proba is None or self.max_proba is None: print("You forgot to initialize the min and max proba") - self._mat = np.zeros((self.number_of_nodes(), self.number_of_nodes())) + self._mat = np.zeros((self.number_of_nodes(), + self.number_of_nodes())) else: self._mat = self.max_proba * np.random.rand(len(self), len(self) ) * np.asarray(nx.adjacency_matrix(self).todense().T) @@ -73,7 +74,8 @@ class InfluenceGraph(nx.DiGraph): print("adding sparse edges to the graph") #Adding sparse non-edges to the graph! self._mat += .1 * np.random.rand(len(self), len(self) - ) * np.random.binomial(1, p=.33, size=(len(self), len(self))) + ) * np.random.binomial(1, p=.33, size=(len(self), + len(self))) return self._mat @property diff --git a/src/convex_optimization.py b/src/convex_optimization.py index 0d506e1..000143f 100644 --- a/src/convex_optimization.py +++ b/src/convex_optimization.py @@ -33,7 +33,7 @@ def sparse_recovery(lbda, n_cascades): y = lbda * theta_.norm(1) - 1./n_cascades*( (w).dot(tensor.log(1-tensor.exp(M.dot(theta_)))) + (1-w).dot(tensor.dot(M, theta_)) - ) + ) z = tensor.row().T z_ = z.flatten() @@ -134,7 +134,7 @@ def test(): A = cascade_creation.generate_cascades(G, .1, 500) M_val, w_val = cascade_creation.icc_matrixvector_for_node(A, 2) print(len(M_val)) - + #Type lasso if 0: f_x, f_xz = type_lasso(lbda) diff --git a/src/make_plots.py b/src/make_plots.py index c45479e..8970cdf 100644 --- a/src/make_plots.py +++ b/src/make_plots.py @@ -54,7 +54,7 @@ def plot_watts_strogatz_graph(): plt.clf() fig = plt.figure(1) labels = [50, 100, 500, 1000, 2000, 5000] - x = [np.log(50), np.log(100), np.log(500), + x = [np.log(50), np.log(100), np.log(500), np.log(1000), np.log(2000), np.log(5000)] sparse_recov = [.25, .32, .7, .82, .89, .92] max_likel = [.21, .29, .67, .8, .87, .9] @@ -84,7 +84,7 @@ def plot_barabasi_albert_graph(): plt.clf() fig = plt.figure(1) labels = [50, 100, 500, 1000, 2000, 5000] - x = [np.log(50), np.log(100), np.log(500), + x = [np.log(50), np.log(100), np.log(500), np.log(1000), np.log(2000), np.log(5000)] sparse_recov = [.35, .38, .58, .69, .79, .86] max_likel = [.35, .38, .56, .68, .78, .85] @@ -188,13 +188,13 @@ def plot_ROC_curve(figure_name): plt.ylabel("Precision") plt.grid(color="lightgrey") - ax.plot(recall_lasso_200, precision_lasso_200, 'ko-', + ax.plot(recall_lasso_200, precision_lasso_200, 'ko-', color="lightseagreen", label="Lasso-200 cascades") - ax.plot(recall_sparse_200, precision_sparse_200, 'ko-', + ax.plot(recall_sparse_200, precision_sparse_200, 'ko-', color="k", label="Our Method-200 cascades") - ax.plot(recall_lasso_50, precision_lasso_50, 'ko-', + ax.plot(recall_lasso_50, precision_lasso_50, 'ko-', color="orange", label="Lasso-50 cascades") - ax.plot(recall_sparse_50, precision_sparse_50, 'ko-', + ax.plot(recall_sparse_50, precision_sparse_50, 'ko-', color="cornflowerblue", label="Our Method-50 cascades") plt.legend(loc="upper right", fontsize=14) plt.savefig("../paper/figures/"+"ROC_curve.pdf") @@ -253,9 +253,9 @@ if __name__=="__main__": #algorithms.greedy_prediction) #convex_optimization.type_lasso) if 0: - compute_graph("../datasets/kronecker_graph_256_cross.txt", + compute_graph("../datasets/kronecker_graph_256_cross.txt", n_cascades=50, lbda=0., min_proba=.2, max_proba=.7, passed_function= - convex_optimization.sparse_recovery, + convex_optimization.sparse_recovery, #convex_optimization.type_lasso, - sparse_edges=True)
\ No newline at end of file + sparse_edges=True) |