时间戳

首页 > 解决方案 > 具有最小长度的序列

python - 具有最小长度的序列

问题描述

(73, 82)如果序列具有最小长度，则应标记位于两个阈值之间的值序列(12)。

该功能sequence是我正在寻找的：

df['between']  = df[df['preasure'].between(73, 82)]
df['sequence'] = sequence(df.between, 12)

标签： pythonpandasnumpy

解决方案

这是一个 numpy 解决方案，我确信可以将其翻译为 pandas（作为读者练习）。

low = 73
high = 82
min_len = 12

首先屏蔽所有位置：

mask = ((x >= low) & (x <= high))

然后找到掩码更改符号的位置：

swap = np.diff(np.r_[False, mask, False])

额外的填充是为了正确表示第一个和最后一个元素。现在将其转换为索引：

ind = np.flatnonzero(swap)

的每个偶数元素都ind代表运行的开始。奇数元素是终点。段的长度由下式给出

start = ind[::2]
end = ind[1::2]
lens = end - start

您可以根据长度选择要选择或删除的索引：

lmask = lens >= min_length

要从索引制作掩码，请使用np.cumsum. 在这里，我们基本上将原始掩码替换为阈值掩码：

result = np.zeros(x.shape, dtype=np.int8)
result[start[lmask]] = +1
lmask[-1] = False  # This eliminates the element past the end of the array
result[end[lmask]] = -1
result = np.cumsum(result, out=result).view(bool)

例子

这是一个小例子，向您展示每一步都发生了什么

low = 3
high = 7
n = 3

np.random.seed(1)
x = np.random.randint(low - 2, high + 2, 20)
# 6, 9, 6, 1, 1, 2, 8, 7, 3, 5, 6, 3, 5, 3, 5, 8, 8, 2, 8, 1
mask = (x >= low) & (x <= high)
# 1, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 1, 1, 1, 0, 1, 1

swap = np.diff(np.r_[False, mask, False])
# 1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1
ind = np.flatnonzero(swap)
# 0,       3,       6,    8,         12,13,14,   17,   18,   20

start = ind[::2]
# 0, 6, 12, 14, 18
end = ind[1::2]
# 3, 8, 13, 17, 20
lens = end - start
# 3, 2, 1, 3, 2
lmask = lens >= min_length
# 1, 0, 0, 1, 0

result = np.zeros(x.shape, dtype=np.int8)
# 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
result[start[lmask]] = +1
# 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0
lmask[-1] = False  # This eliminates the element past the end of the array
result[end[lmask]] = -1
# 1, 0, 0,-1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,-1, 0, 0
result = np.cumsum(result, out=result).view(bool)
# 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0

TL;博士

def sequence(x, low, high, thresh):
    ind = np.flatnonzero(np.diff(np.r_[False, ((x >= low) & (x <= high)), False]))
    start = ind[::2]
    end = ind[1::2]
    mask = end - start >= thresh
    result = np.zeros(x.size, dtype=np.int8)
    result[start[mask]] = +1
    mask[-1] = False
    result[end[mask]] = -1
    return np.cumsum(result, out=result).view(bool)

推荐阅读