[英]python vectorization: how to increase the efficiency of 4 layer loop
我正在嘗試使用 Gibbs 采樣實現 LDA,在更新每個主題比例的步驟中,我有一個 4 層循環,它運行非常慢,我不知道如何提高這段代碼的效率。 我現在擁有的代碼如下:
N_W 是單詞數,N_D 是文檔數,Z[i,j] 是主題分配(1 到 K 個可能的分配),X[i,j] 是第 j 個單詞的計數第 i 個文檔 Beta[k,:] 的維度為 [K, N_W]。
更新如下:
for k in range(K): # iteratively for each topic update
n_k = np.zeros(N_W) # vocab size
for w in range(N_W):
for i in range(N_D):
for j in range(N_W):
# counting number of times a word is assigned to a topic
n_k[w] += (X[i,j] == w) and (Z[i,j] == k)
# update
Beta[k,:] = np.random.dirichlet(gamma + n_k)
您可以使用邏輯函數擺脫最后兩個 for 循環:
for k in range(K): # iteratively for each topic update
n_k = np.zeros(N_W) # vocab size
for w in range(N_W):
a = np.logical_not(X-w) # all X(i,j) == w become a True, others a false
b = np.logical_not(Z-k) # all Z(i,j) == w become a True, others a false
c = np.logical_and(a,b) # all (i,j) where X(i,j) == w and Z(i,j) == k are True, others false
n_k[w] = np.sum(c) # sum all True values
甚至作為一個班輪:
n_k = np.array([[np.sum(np.logical_and(np.logical_not(X[:N_D,:N_W]-w), np.logical_not(Z[:N_D,:N_W]-k))) for w in range(N_W)] for k in range(K)])
然后 n_k 中的每一行都可以用於 beta 計算。 現在它還包括 N_W 和 N_D 作為限制,如果它們不等於 X 和 Z 的大小
我對以下矩陣進行了一些測試:
import numpy as np
K = 90
N_W = 100
N_D = 11
N_W = 12
Z = np.random.randint(0, K, size=(N_D, N_W))
X = np.random.randint(0, N_W, size=(N_D, N_W))
gamma = 1
原代碼:
%%timeit
Beta = numpy.zeros((K, N_W))
for k in range(K): # iteratively for each topic update
n_k = np.zeros(N_W) # vocab size
for w in range(N_W):
for i in range(N_D):
for j in range(N_W):
# counting number of times a word is assigned to a topic
n_k[w] += (X[i,j] == w) and (Z[i,j] == k)
# update
Beta[k,:] = np.random.dirichlet(gamma + n_k)
865 ms ± 8.37 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
然后僅對內部兩個循環進行矢量化:
%%timeit
Beta = numpy.zeros((K, N_W))
for k in range(K): # iteratively for each topic update
n_k = np.zeros(N_W) # vocab size
for w in range(N_W):
n_k[w] = np.sum((X == w) & (Z == k))
# update
Beta[k,:] = np.random.dirichlet(gamma + n_k)
21.6 ms ± 542 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)
最后是一些廣播和提取常見元素的創造性應用:
%%timeit
Beta = numpy.zeros((K, N_W))
w = np.arange(N_W)
X_eq_w = np.equal.outer(X, w)
for k in range(K): # iteratively for each topic update
n_k = np.sum(X_eq_w & (Z == k)[:, :, None], axis=(0, 1))
# update
Beta[k,:] = np.random.dirichlet(gamma + n_k)
4.6 ms ± 92.1 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)
這里的權衡是在速度和 memory 之間。 對於我使用的形狀,這不是內存密集型,但我在最后一個解決方案中構建的中間三維 arrays 可能會變得相當大。
聲明:本站的技術帖子網頁,遵循CC BY-SA 4.0協議,如果您需要轉載,請注明本站網址或者原文地址。任何問題請咨詢:yoyou2525@163.com.