问题:

I have two 2D number arrays i.e., lat and lon each of (95,60) dimensions. I am trying to find the indices for the following conditional statement.

ind = np.where(((lat >= 20.0 and lat <= 30.0) and (lon >= 90.0 and lon <= 100.0)))

I get the following error:

ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

My lat array contains the data like this:

array([[ 19.13272095,  19.52386856,  19.8377533 , ...,  22.81465149,          22.80446053,  22.78530312],        [ 19.2490139 ,  19.64085197,  19.95526123, ...,  22.93462563,          22.92414474,  22.90457916],        [ 19.36528587,  19.75781822,  20.07275009, ...,  23.05456543,          23.04379272,  23.02381706],        ...,        [ 29.7395134 ,  30.20263481,  30.57126617, ...,  33.85147858,          33.81478882,  33.75755692],        [ 29.85359764,  30.31761169,  30.68692398, ...,  33.97143555,          33.93445587,  33.87679672],        [ 29.9676609 ,  30.4325695 ,  30.80256653, ...,  34.09137726,          34.05410767,  33.99602509]], dtype=float32)

Similarly, lon array contains the data as follows:

array([[  96.78914642,   98.04283905,   99.0661087 , ...,  119.39452362,          120.45418549,  121.74817657],        [  96.75196075,   98.00632477,   99.03016663, ...,  119.37508392,          120.43569183,  121.73084259],        [  96.71466064,   97.96970367,   98.99414062, ...,  119.35567474,          120.41724396,  121.71356201],        ...,        [  92.90063477,   94.2386322 ,   95.33486176, ...,  117.72390747,          118.90248108,  120.34107971],        [  92.85238647,   94.19155884,   95.2888031 , ...,  117.7070694 ,          118.88733673,  120.32800293],        [  92.80393219,   94.14429474,   95.24256897, ...,  117.69021606,          118.87218475,  120.31491089]], dtype=float32)

I tried to revise my command as follows:

ind = np.where(((lat.all() >= 20.0 and lat.all() <= 30.0) and (lon.all() >= 90.0 and lon.all() <= 100.0)))

I get the following empty array as output:

(array([], dtype=int64),)

I have even tried the following:

ind = np.where(lat.all() >= 20.0 and lat.all() <= 30.0)

Output is obtained as: (array([], dtype=int64),)

I don't understand how to solve my problem. Can anyone help me how and where I am doing wrong?

回答1:

This needs two fixes, () to perform the comparisons first, and replacing the scalar and with an elementwise &

ind = np.where((((lat >= 20.0) & (lat <= 30.0)) & ((lon >= 90.0) & (lon <= 100.0))))

This ValueError arises when an boolean array is used in a context that needs a scalar boolean. if lat >= 20.0: is common case. and is a Python short circuiting operator, so in effect does a if test.

We also need to do the (<=) tests before the &

With just part of the lat array:

Out[48]: (array([], dtype=int32), array([], dtype=int32)) In [49]: np.where((((lat >= 20.0) & (lat <= 30.0)))) Out[49]:  (array([0, 0, 0, 1, 1, 1, 2, 2, 2, 2, 3, 4, 5], dtype=int32),  array([3, 4, 5, 3, 4, 5, 2, 3, 4, 5, 0, 0, 0], dtype=int32))

回答2:

Never liked a big pile of parentheses and ampersands. I usually use functools.reduce in this situation.

from functools import reduce  ind = np.where(reduce(np.logical_and , (lat >= 20.0 , lat <= 30.0 , lon >= 90.0 , lon <= 100.0)))

转载请标明出处:Find indices of 2D arrays using conditional statement in Python

文章来源: Find indices of 2D arrays using conditional statement in Python

标签

array