Skip to content

Instantly share code, notes, and snippets.

View ducnh1022's full-sized avatar

Nguyen Hong Duc ducnh1022

1 follower · 4 following
View GitHub Profile
@ducnh1022
ducnh1022 / gist:3ebf826936daf031c1d9ac034311ae62
Created March 28, 2026 05:04
-- handle string float, numeric, integer
SAFE_CAST(
REGEXP_REPLACE(col, r"[^0-9eE+\\-\\.]", "")
AS FLOAT64)
-- handle string boolean
CASE
WHEN col IS NULL THEN NULL
WHEN LOWER(TRIM(col)) IN ("true", "1", "yes", "y", "t") THEN TRUE
WHEN LOWER(TRIM(col)) IN ("false", "0", "no", "n", "f") THEN FALSE
@ducnh1022
ducnh1022 / gist:9972ae70b31f81d7fa31b6c23ab634f3
Created March 25, 2026 19:37
def get_table_info(project_id, dataset_id, table_id, sample_size=5):
table_ref = f"{project_id}.{dataset_id}.{table_id}"
# Get schema
table = client.get_table(table_ref)
schema = table.schema
# Get sample data
query = f"""
SELECT *
@ducnh1022
ducnh1022 / gist:606a1821836e14de6555dda98a31bf9c
Created March 5, 2026 17:47
import difflib
def fuzzy_insert(words, new_word, threshold=0.6):
# 1️⃣ Calculate similarity scores
similarities = [
(i, difflib.SequenceMatcher(None, new_word, w).ratio())
for i, w in enumerate(words)
]
# 2️⃣ Find best match
@ducnh1022
ducnh1022 / gist:701b48cc874a2a617d227ef7b4697413
Created February 24, 2026 10:46
import torch
import torchaudio
import numpy as np
import soundfile as sf
from speechbrain.pretrained import SpeakerRecognition
from scipy.spatial.distance import cosine
# ============================
# CONFIG
@ducnh1022
ducnh1022 / gist:72585412f4b3dd06da5a895da2add58d
Created February 24, 2026 06:45
import sounddevice as sd
import soundfile as sf
import numpy as np
import queue
import datetime
import os
# ==============================
# CONFIG
# ==============================
@ducnh1022
ducnh1022 / gist:06355de13df1975f8decbc57951f1b9e
Created February 24, 2026 06:44
import webrtcvad
import collections
import sys
import wave
import contextlib
from pydub import AudioSegment
def read_wave(path):
with contextlib.closing(wave.open(path, 'rb')) as wf:
@ducnh1022
ducnh1022 / gist:51848345be0d4defcb65b40ddbb76e1c
Created February 24, 2026 05:45
import collections
import sys
import wave
import time
import webrtcvad
import pyaudio
# Audio settings
FORMAT = pyaudio.paInt16
CHANNELS = 1
@ducnh1022
ducnh1022 / gist:23d2f1f70c221bed4104287b99732e8e
Created February 23, 2026 08:05
# Most compatible way – use .loc / .iat if DataFrame, indexing if ndarray
loudness = float(features['loudness_sma3_amean'].item()
if hasattr(features, '__getitem__') and 'loudness_sma3_amean' in features
else features[smile.feature_names.index('loudness_sma3_amean')])
voicing_key = 'voicingProb_sma3nz_amean'
voicing = 0.0
if voicing_key in smile.feature_names:
idx = smile.feature_names.index(voicing_key)
voicing = float(features[voicing_key].item()
@ducnh1022
ducnh1022 / gist:2481f4acbf53d737cffc09298dbec338
Created February 23, 2026 07:47
import pyaudio
import numpy as np
import opensmile
import time
import datetime
import collections
import soundfile as sf # pip install soundfile
# ================= Configuration =================
SAMPLE_RATE = 16000
@ducnh1022
ducnh1022 / game_hanzi.html
Created February 8, 2026 04:12
game hanzi
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>VerbVibe - Trilingual Master</title>
<link href="https://fonts.googleapis.com/css2?family=Montserrat:wght@400;700&display=swap" rel="stylesheet">
<style>
:root {
--primary-blue: #A2D2FF;