my-asu-archive/CSE515_MWDB_Project
1
0
Fork 0
mirror of https://github.com/20kaushik02/CSE515_MWDB_Project.git synced 2025-12-06 09:24:07 +00:00
Code Issues Packages Projects Releases Wiki Activity

lsh small uupdate

Browse Source
This commit is contained in:
pranavbrkr 2023-11-28 15:46:47 -07:00
parent b68007f647
commit b6ff52843f
2 changed files with 30 additions and 37 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

55
Phase 3/task_4.ipynb
View File

File diff suppressed because one or more lines are too long

12
Phase 3/utils.py
View File

@ -175,7 +175,7 @@ class LSH:
self.num_hashes = num_hashes self.num_hashes = num_hashes
self.hash_tables = [defaultdict(list) for _ in range(num_layers)] self.hash_tables = [defaultdict(list) for _ in range(num_layers)]
self.unique_images_considered = set() self.unique_images_considered = set()
self.overall_images_considered = set() self.overall_images_considered = []
self.create_hash_tables() self.create_hash_tables()
def hash_vector(self, vector, seed): def hash_vector(self, vector, seed):
@ -192,7 +192,7 @@ class LSH:
def find_similar(self, external_image, t): def find_similar(self, external_image, t):
similar_images = set() similar_images = set()
visited_buckets = set() visited_buckets = set()
unique_images_considered = set() unique_images_considered = []
for layer in range(self.num_layers): for layer in range(self.num_layers):
hash_code = self.hash_vector(external_image, seed=layer) hash_code = self.hash_vector(external_image, seed=layer)
@ -202,7 +202,7 @@ class LSH:
if hash_code in self.hash_tables[layer]: if hash_code in self.hash_tables[layer]:
for idx in self.hash_tables[layer][hash_code]: for idx in self.hash_tables[layer][hash_code]:
similar_images.add(idx) similar_images.add(idx)
unique_images_considered.add(idx) unique_images_considered.append(idx)
# Searching in nearby buckets based on Hamming distance # Searching in nearby buckets based on Hamming distance
for key in self.hash_tables[layer]: for key in self.hash_tables[layer]:
@ -211,10 +211,10 @@ class LSH:
for idx in self.hash_tables[layer][key]: for idx in self.hash_tables[layer][key]:
similar_images.add(idx) similar_images.add(idx)
unique_images_considered.add(idx) unique_images_considered.append(idx)
self.unique_images_considered = unique_images_considered self.overall_images_considered = unique_images_considered
self.overall_images_considered = similar_images self.unique_images_considered = set(unique_images_considered)
similarities = [ similarities = [
(idx, self.euclidean_distance(external_image, self.data[idx])) for idx in similar_images (idx, self.euclidean_distance(external_image, self.data[idx])) for idx in similar_images