Filter array of values by their two first characters

user_8551424 0
Viewed 213

I have an array of values and I want to filter it by their two first characters.

Now my values are strings that always start with an identifier of two characters like 01, 02, 03... Thus I'm doing the following:

payloads = []
for pay in all_payloads:
    if pay[:2] == '01':
        payloads.append(pay)

I know I can do a list comprehension too, but actually what I want to learn if its possible to filter it directly with numpy. I thought maybe np.where() it's a good option but it's not working for me...

cosl = np.where(all_payloads[:2] == '01')

Can somebody help me to learn how to do it?

Sample inputs

['011b009f2e5141cc3c07be7756e0418a9ce6bc3e1ab13e9e1ffc3f0418943e00000000'
 '011c00f51a51411de008be6f0fe541baa256bd85c2be3eb24af13fdbcb833e00000000'
 '031d0086265141240401be873fea414b64e93b4d0a923e0723f33f2c95a63e00000000'
 '011e00e62c5141a3e306be12c2ef4135d44a3c7b0e773eff51fb3f67da883e00000000'
 '021e00e62c5141a3e306be12c2ef4135d44a3c7b0e773eff51fb3f67da883e00000000'
 '011f0035335141f57107be6806f6410664e9bcd088983ee52ef13fb4b48b3e00000000'
 '012000af405141521704be0184fd41d07752bd5eadb13ef0adf33fe4279c3e00000000'
 '032100ae4351415a94fbbdf9e8024296710b3e4417c93fd77efa3f35b1043f00000000'
 '01220015525141c20601be130e06423648f7bc9341a63e6c35fd3f6972943e00000000']

Thank you very much!

numpy python
2 Answers

You can use functions from np.char module for doing this:

a = np.array(["01_a", "01_b", "01_c", "02_a", "02_b", "02_c", "03_a", "03_b", "03_c"])
indices = np.where(np.char.startswith(a, "01"))
print(a[indices])

Output:

['01_a' '01_b' '01_c']

Edit:

Output for your added sample input for '01' filtering:

['011b009f2e5141cc3c07be7756e0418a9ce6bc3e1ab13e9e1ffc3f0418943e00000000'
'011c00f51a51411de008be6f0fe541baa256bd85c2be3eb24af13fdbcb833e00000000'
'011e00e62c5141a3e306be12c2ef4135d44a3c7b0e773eff51fb3f67da883e00000000'
'011f0035335141f57107be6806f6410664e9bcd088983ee52ef13fb4b48b3e00000000'
'012000af405141521704be0184fd41d07752bd5eadb13ef0adf33fe4279c3e00000000'
'01220015525141c20601be130e06423648f7bc9341a63e6c35fd3f6972943e00000000']

indices:

(array([0, 1, 3, 5, 6, 8], dtype=int64),)
user_160097040

You can also do filtering using the builtin functions. My take on it is:

filtered = list(filter(lambda x: x.startswith('01'), inputs))
user_105007570
Related