Adv indexing

[rsokl]

@davidmascharka can you proof this?

[davidmascharka]

Looks good overall. Here's what I've got:

L66: This is probably not worth anything, but your callout box says "Definition: Advanced Indexing" and then doesn't actually define anything. It just describes cases in which your indexing is advanced.

L91: "See that we can acces its an arbitrary number of the arrays contents" -- remove the word its

L105: "The specification for accessing" sounds a little weird to me. I think of a specification more as a spec sheet. Maybe the instruction or something?

L141: you run np.shares_memory(y, y[index]) but don't have an output. Following that should be a line False

L231: "we specified a single index-array array" -- get rid of one of the arrays
L231: "we must specify N index-arrays, one for each dimension" -- replace the comma with a semicolon

L233: "The corresponding entries of each of the N index arrays is used" should be are used

L246: "We specify three index-arrays, the indices to be..." replace that comma with a colon: "three index-arrays: the indices..."

L252: "The other index arrays are formed accordingly." -- I would say "formed similarly" instead

L383: "...by supplying N index-arrays of integers, one for each axis of data." -- replace the comma with a semicolon

L392: Up to you, but Boolean Array Indexing can be read as Boolean-Array Indexing (indexing into boolean arrays) or Boolean Array-Indexing (using booleans for array indexing). I'd hyphenate where you want the reader to read a hyphen in (Boolean Array-Indexing).

L393: "...permits the use of a boolean-value array" should be "boolean-valued"

L418: "...and ind is a boolean-value of the same shape as x." -- boolean values don't have a shape: they're just values. Change to boolean-valued array of the same shape

L437: "Use boolean array indexing NumPy's logical functions" is a mess. Here's a rewrite (also changes floating point to floating-point and changes a , to a ;):

Use boolean array-indexing and NumPy's [logical functions](https://docs.scipy.org/doc/numpy/reference/routines.logic.html) to select the contents of `h` that satisfy the following conditions. Because you are dealing with floating-point numbers, [you should not require that two values are exactly equal](https://www.pythonlikeyoumeanit.com/Module2_EssentialsOfPython/Basic_Objects.html#Understanding-Numerical-Precision); rather, make use of the function [numpy.isclose](https://docs.scipy.org/doc/numpy-dev/reference/generated/numpy.isclose.html). 

L450: "Converting a Boolean Index Array to Integer Index Arrays" should be Boolean Index-Array (hyphenated)

L451: "...can be used to take a boolean-value array" should be "boolean-valued"

L468: "...a tuple of three integer-value index..." should be "integer-valued"
L468: "...was created to access the positive entries of x, the tuple of..." -- comma should be a semicolon

L471: not necessary, but split the line where that comma is and just get rid of the comma:

# unpacking the arrays is not necessary # you can use the tuple as an index >>> ind0, ind1, ind2 = np.where(bool_ind) >>> x[ind0, ind1, ind2]

At this same point, it might be worth just showing that you don't have to unpack by adding this line:

>>> x[np.where(bool_ind)]

L478: "where specific conditions were met" should be "are met"

L494: "boolean array indexing" should be "boolean array-indexing"
L494: "...and the the layout of its contents" -- remove the extra the

L562: "Recall that redundant entries in array" -- "redundant entries in an array"

L603: Not important, but I would say Combining Basic and Advanced Indexing instread of Advanced and Basic

L604: "Integer and boolean-valued arrays" should be "Integer- and boolean-valued arrays" -- note the hyphen on integer

L607: again, not important but I would swap the order on advanced and basic

Run a consistency check for index-array and index array -- there are a few in the text that I didn't mention because the writing slips in and out between the two; pick one and make them all use it.

[davidmascharka]

^ Up there

[rsokl]

@davidmascharka thanks so much for this! 0.11 will be released in a few minutes