added some basics about numpy arrays

sdevikar · sdevikar · commit 938f6d9fb491 · 2017-03-02T17:27:18.000-05:00
diff --git a/python2/13_numpy_basics.py b/python2/13_numpy_basics.py
@@ -0,0 +1,60 @@
+import numpy as np
+
+a = np.arange(15).reshape(3, 5)
+print(a)
+
+###
+# This will create the following array:
+# [
+#  [ 0  1  2  3  4]
+#  [ 5  6  7  8  9]
+#  [10 11 12 13 14]
+# ]
+#
+# Notice that we have 3 arrays of length 5 each.
+# Read the array as follows:
+# The last value in the shape tuple is the length of the innermost array
+# The value to the left in the tuple i.e. 3 in (3,5) is the number of such arrays
+###
+
+b = np.arange(30).reshape(3, 5, 2)
+print(b)
+
+###
+# This will create the following array
+# [
+#  [[ 0  1]
+#  [ 2  3]
+#  [ 4  5]
+#  [ 6  7]
+#  [ 8  9]]
+#
+# [[10 11]
+#  [12 13]
+#  [14 15]
+#  [16 17]
+#  [18 19]]
+#
+# [[20 21]
+#  [22 23]
+#  [24 25]
+#  [26 27]
+#  [28 29]]
+# ]
+# Notice that the reading method above still holds. i.e. in shape tuple (3,5,2)
+# 2 is the length of the innermost array
+# The value to its left, i.e. 5, is the number of the innermost array
+# The value further to the left, i.e. 3, is the number of arrays of arrays of length 2
+
+
+# Indexing the arrays
+print(a[0]) # this should return the first row i.e. [0 1 2 3 4]
+print(a[0:2]) # this should return row 0 and row 1. Row 2 is not included
+
+# in order to specify an individual element, row and column indexes,
+# which are called axes in numpy arrays, should be specified as comma separated values
+print("a[0,1] = {}".format(a[0,1])) # this should return 1
+print("a[0,1] = {}".format(a[0][1])) # this is the same as a[0,1]
+
+# So, in order to access 15 in array b, we can do the following
+print("b[1,2,1] = {}".format(b[1,2,1]))
