仅当在X个其他子列表中存在时保留嵌套列表中的项

collections.Counter 在扁平列表中,使用集合忽略单个子列表中的重复值:

appearances = Counter(word for sub in arr for word in set(sub))
# Counter({'hola': 4, 'hello': 3, 'ciao': 3, 'bonjour': 2, 'namaste': 2})

[[word for word in sub if appearances[word] >= threshold] for sub in arr]

from collections import Counter

def threshold_filter(arr, threshold):
  appearances = Counter(word for sub in arr for word in set(sub))

  return [
    [word for word in sub if appearances[word] >= threshold] 
    for sub in arr
  ]

print(threshold_filter(test, 3))

# Result 
[['hello', 'hola'], ['hello', 'hola'], ['hello', 'ciao'], ['hola', 'ciao'], ['hola', 'ciao'], []]