Aldo ortega

.gitignore

@@ -14,3 +14,6 @@ __pycache__/
 
 # CodSpeed
 .codspeed/
+
+# Mine
+codes

.python-version

@@ -1 +1 @@
-3.15t
+3.14

rounds/1_histogram/solution.py

@@ -4,11 +4,28 @@
 passes out of the box. Replace the body of ``compute_histogram`` with your
 own faster implementation.
 """
+import numpy as np
 
 
 def compute_histogram(path: str) -> dict[bytes, int]:
     """Frequency of every 2-byte bigram in the file at ``path``."""
-    # TODO: remove this delegation and write your own implementation here.
-    from .baseline import compute_histogram as _baseline
+    # Step 1: read the whole file into memory as a single bytes object.
+    with open(path, "rb") as f:
+        data = f.read()
 
-    return _baseline(path)
+    # Step 2: slide a 2-byte window across the buffer. For ``b"ABCD"`` the
+    # iterations produce ``b"AB"``, ``b"BC"``, then ``b"CD"``. For each window,
+    # bump the matching bucket in a ``dict`` keyed by the bigram itself.
+    counts= [[0] * 256 for _ in range(256)]
+
+    for i in range(len(data) - 1):
+        a, b = data[i], data[i + 1]
+        counts[a][b] += 1
+
+    result = {}
+    for i, row in enumerate(counts):
+        for j, count in enumerate(row):
+            if count > 0:
+                bigram = bytes([i, j])
+                result[bigram] = count
+    return result

rounds/3_dna/solution.py

@@ -6,12 +6,57 @@
 """
 
 from .baseline import find_matches as _baseline
+from threading import Thread
+import numpy as np
 
 
 def find_matches(fasta_path: str, pattern: bytes) -> list[tuple[str, list[int]]]:
     """Find every FASTA record whose sequence contains ``pattern``.
 
     Returns ``[(record_id, [positions...]), ...]`` in file order.
     """
-    # TODO: remove this delegation and write your own implementation here.
-    return _baseline(fasta_path, pattern)
+    # Step 1: read the whole FASTA file as text and decode the pattern so the
+    # search below can use a single ``str`` API.
+    pattern_str = pattern.decode("ascii")
+    with open(fasta_path, "rb") as f:
+        text = f.read()
+    matches: list[tuple[str, list[int]]] = []
+
+    # Step 2: split the file on '>' to peel off one record at a time. The
+    # first element is the chunk before any header (empty for well-formed
+    # files) and is skipped by the ``.strip()`` guard below.
+    sequences = []
+    for record in text.split(b">"):
+        if not record.strip():
+            continue
+
+        # Step 3: a record looks like ``"<id>\n<seq line 1>\n<seq line 2>\n..."``.
+        # The id is the first line; the remaining lines are joined back into a
+        # single contiguous sequence string.
+        lines = record.split(b"\n")
+        record_id = lines[0].strip().decode("ascii")
+        sequence = b"".join(lines[1:]).replace(b" ", b"").decode("ascii")
+        sequences.append((record_id, sequence))
+
+    threads = []
+    for record_id, sequence in sequences:
+        thread = Thread(target=match_record, args=(record_id, sequence, pattern_str, matches))
+        thread.start()
+        threads.append(thread)
+    for thread in threads:
+        thread.join()
+
+    return matches
+
+def match_record(record_id, sequence, pattern_str, matches):
+    positions: list[int] = []
+    start = 0
+    while True:
+        pos = sequence.find(pattern_str, start)
+        if pos == -1:
+            break
+        positions.append(pos)
+        start = pos + 1
+
+    if positions:
+        matches.append((record_id, positions))