Introduction to NumPy and Matplotlib

Ch2-6 NumPy Arrays

More efficient
- Use for large scale arrays / matrices
Convert between lists
tolist() and np.array()
.shape check the dimensions
.dtype check the data type of the elements
np.zeros() creates an array filled with 0s of the set shape
np.arange(n) creates an array from 0 to n-1
np.linspace(s, e, num=g) creates an array with elements, from s to e linearly separated with gap size g
+, -, +=, and -= work elementwise
Boolean indexing
- arr > n Maps the array to booleans, compares them elementwise
- arr[arr > n] keeps all elements that satisfy the condition
Array stacking
- np.vstack() - stack on vertical stack
- np.hstack() - stack on horizontal stack
- np.dstack() - stack on depth stack
Array concatenation
- Appends on the same axis, unlike stacking
Array splitting
- Can specify splitting at what positions
- np.vsplit() - splits on vertical axis
- np.hsplit() - splits on horizontal axis
- np.dsplit() - splits on depth axis
Repeating arrays
- np.repeat() repeats items elementwise
  - Flattens array if no axis is specified
Nonzero searching
- np.nonzero() returns the indices in arrays
Unique filtering
- np.unique() returns the unique elements
  - Makes a set
Sorting
- arr.sort() is a mutator
- np.argsort returns the sorted indices

Saving and loading
- Will save as .npz
- Can also save as txt by .savetxt() and .loadtxt()
Random
- Seeding
  - rg = default_rng(seed)
- Use rg for functions
  - .random(), .integers()
Print options
- np.set_printoptions(threshold=x)
  - Prints more items up to x