The rfind() function of the numpy.strings module

Overview:

For each string element found in a NumPy array-like the the function numpy.strings.rfind() function finds the highest index at which a substring is available. The indexes are returned as an ndarray of integers.
The substring can be searched between a start index and stop index. When start and stop indexes are not specified the string is searched from the position 0 to length – 1.
If the input parameter is a scalar, the rfind() function returns an integer denoting the highest index position at which a substring is found in the string.
The function works for unicode strings (numpy.str_), byte strings(numpy.bytes_) and variable width strings(numpy.dtypes.StrDType).
When a NumPy array is created with elements having type as Python string NumPy converts it into a NumPy array of numpy.str_ type.

Example:

# Example Python program that finds
# substrings on the string elements of the NumPy array
# using numpy.strings.rfind() function
import numpy

# Create an ndarray containing texts of variable length
lines = numpy.ndarray(shape = (1, 3), dtype = numpy.dtypes.StringDType())

lines[0][0] = "Rose is a rose is a rose is a rose."
lines[0][1] = "Jack Rose Jack Rose."
lines[0][2] = "It is rose in hen."

# Find the substring
substr = "Rose"
rIndexes = numpy.strings.rfind(lines, substr)
print("Text lines:")
print(lines)
print("Substr \"%s\" found in indexes:"%substr)
print(rIndexes)

Output:

Text lines:

[['Rose is a rose is a rose is a rose.' 'Jack Rose Jack Rose.'

'It is rose in hen.']]

Substr "Rose" found in indexes:

[[ 0 15 -1]]