From 56672c1a1f65cb8917d639de5085273b81761e0b Mon Sep 17 00:00:00 2001
From: Andrei Ivanov <andreii@nvidia.com>
Date: Thu, 30 Jan 2025 16:14:30 -0800
Subject: [PATCH 1/2] Addressing warnings related to the `multi_class`
parameter in `LogisticRegression`.
---
.../distributed/graphsage/node_classification_unsupervised.py | 2 +-
examples/pytorch/graphsage/advanced/model.py | 2 +-
examples/pytorch/graphsage/dist/train_dist_unsupervised.py | 2 +-
examples/pytorch/metapath2vec/test.py | 4 ++--
examples/pytorch/multigpu/multi_gpu_link_prediction.py | 2 +-
examples/pytorch/node2vec/model.py | 2 +-
6 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/examples/distributed/graphsage/node_classification_unsupervised.py b/examples/distributed/graphsage/node_classification_unsupervised.py
index 4b71f599064b..2b030646bf28 100644
--- a/examples/distributed/graphsage/node_classification_unsupervised.py
+++ b/examples/distributed/graphsage/node_classification_unsupervised.py
@@ -175,7 +175,7 @@ def compute_acc(emb, labels, train_nids, val_nids, test_nids):
labels = labels.cpu().numpy()
emb = (emb - emb.mean(0, keepdims=True)) / emb.std(0, keepdims=True)
- lr = lm.LogisticRegression(multi_class="multinomial", max_iter=10000)
+ lr = lm.LogisticRegression(max_iter=10000)
lr.fit(emb[train_nids], labels[train_nids])
pred = lr.predict(emb)
diff --git a/examples/pytorch/graphsage/advanced/model.py b/examples/pytorch/graphsage/advanced/model.py
index 7dede9643cc4..edb965b78936 100644
--- a/examples/pytorch/graphsage/advanced/model.py
+++ b/examples/pytorch/graphsage/advanced/model.py
@@ -104,7 +104,7 @@ def compute_acc_unsupervised(emb, labels, train_nids, val_nids, test_nids):
emb = (emb - emb.mean(0, keepdims=True)) / emb.std(0, keepdims=True)
- lr = lm.LogisticRegression(multi_class="multinomial", max_iter=10000)
+ lr = lm.LogisticRegression(max_iter=10000)
lr.fit(emb[train_nids], train_labels)
pred = lr.predict(emb)
diff --git a/examples/pytorch/graphsage/dist/train_dist_unsupervised.py b/examples/pytorch/graphsage/dist/train_dist_unsupervised.py
index f3eb1a24305b..24d737f4f280 100644
--- a/examples/pytorch/graphsage/dist/train_dist_unsupervised.py
+++ b/examples/pytorch/graphsage/dist/train_dist_unsupervised.py
@@ -175,7 +175,7 @@ def compute_acc(emb, labels, train_nids, val_nids, test_nids):
labels = labels.cpu().numpy()
emb = (emb - emb.mean(0, keepdims=True)) / emb.std(0, keepdims=True)
- lr = lm.LogisticRegression(multi_class="multinomial", max_iter=10000)
+ lr = lm.LogisticRegression(max_iter=10000)
lr.fit(emb[train_nids], labels[train_nids])
pred = lr.predict(emb)
diff --git a/examples/pytorch/metapath2vec/test.py b/examples/pytorch/metapath2vec/test.py
index 7f293daa1b7f..6ec68f86c7f1 100644
--- a/examples/pytorch/metapath2vec/test.py
+++ b/examples/pytorch/metapath2vec/test.py
@@ -95,11 +95,11 @@
file.close()
print("beging predicting")
clf_venue = LogisticRegression(
- random_state=0, solver="lbfgs", multi_class="multinomial"
+ random_state=0, solver="lbfgs",
).fit(venue_training, venue_label)
y_pred_venue = clf_venue.predict(venue_testing)
clf_author = LogisticRegression(
- random_state=0, solver="lbfgs", multi_class="multinomial"
+ random_state=0, solver="lbfgs",
).fit(author_training, author_label)
y_pred_author = clf_author.predict(author_testing)
macro_average_venue += f1_score(
diff --git a/examples/pytorch/multigpu/multi_gpu_link_prediction.py b/examples/pytorch/multigpu/multi_gpu_link_prediction.py
index bb0c64f75c4f..52b5204e4c8e 100644
--- a/examples/pytorch/multigpu/multi_gpu_link_prediction.py
+++ b/examples/pytorch/multigpu/multi_gpu_link_prediction.py
@@ -142,7 +142,7 @@ def compute_acc_unsupervised(emb, labels, train_nids, val_nids, test_nids):
test_nids = test_nids.cpu().numpy()
test_labels = labels[test_nids]
emb = (emb - emb.mean(0, keepdims=True)) / emb.std(0, keepdims=True)
- lr = lm.LogisticRegression(multi_class="multinomial", max_iter=10000)
+ lr = lm.LogisticRegression(max_iter=10000)
lr.fit(emb[train_nids], train_labels)
pred = lr.predict(emb)
f1_micro_eval = skm.f1_score(val_labels, pred[val_nids], average="micro")
diff --git a/examples/pytorch/node2vec/model.py b/examples/pytorch/node2vec/model.py
index 3b2db93801c0..2dce23225d3e 100644
--- a/examples/pytorch/node2vec/model.py
+++ b/examples/pytorch/node2vec/model.py
@@ -191,7 +191,7 @@ def evaluate(self, x_train, y_train, x_val, y_val):
x_train, y_train = x_train.cpu().numpy(), y_train.cpu().numpy()
x_val, y_val = x_val.cpu().numpy(), y_val.cpu().numpy()
lr = LogisticRegression(
- solver="lbfgs", multi_class="auto", max_iter=150
+ solver="lbfgs", max_iter=150
).fit(x_train, y_train)
return lr.score(x_val, y_val)
From f0e415ef36670ccc9049c5ff581d4ac8ee073a74 Mon Sep 17 00:00:00 2001
From: Andrei Ivanov <andreii@nvidia.com>
Date: Thu, 30 Jan 2025 20:27:42 -0800
Subject: [PATCH 2/2] Fixing lint problems
---
examples/pytorch/metapath2vec/test.py | 12 ++++++------
examples/pytorch/node2vec/model.py | 6 +++---
2 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/examples/pytorch/metapath2vec/test.py b/examples/pytorch/metapath2vec/test.py
index 6ec68f86c7f1..ca6ab0ae268f 100644
--- a/examples/pytorch/metapath2vec/test.py
+++ b/examples/pytorch/metapath2vec/test.py
@@ -94,13 +94,13 @@
author_true = np.array(author_true)
file.close()
print("beging predicting")
- clf_venue = LogisticRegression(
- random_state=0, solver="lbfgs",
- ).fit(venue_training, venue_label)
+ clf_venue = LogisticRegression(random_state=0, solver="lbfgs").fit(
+ venue_training, venue_label
+ )
y_pred_venue = clf_venue.predict(venue_testing)
- clf_author = LogisticRegression(
- random_state=0, solver="lbfgs",
- ).fit(author_training, author_label)
+ clf_author = LogisticRegression(random_state=0, solver="lbfgs").fit(
+ author_training, author_label
+ )
y_pred_author = clf_author.predict(author_testing)
macro_average_venue += f1_score(
venue_true, y_pred_venue, average="macro"
diff --git a/examples/pytorch/node2vec/model.py b/examples/pytorch/node2vec/model.py
index 2dce23225d3e..282799bbe1fb 100644
--- a/examples/pytorch/node2vec/model.py
+++ b/examples/pytorch/node2vec/model.py
@@ -190,9 +190,9 @@ def evaluate(self, x_train, y_train, x_val, y_val):
x_train, y_train = x_train.cpu().numpy(), y_train.cpu().numpy()
x_val, y_val = x_val.cpu().numpy(), y_val.cpu().numpy()
- lr = LogisticRegression(
- solver="lbfgs", max_iter=150
- ).fit(x_train, y_train)
+ lr = LogisticRegression(solver="lbfgs", max_iter=150).fit(
+ x_train, y_train
+ )
return lr.score(x_val, y_val)