# Concatenation And Splitting ## 1. CONCATENATING ARRAYS `np.concatenate()` constructor is used to concatenate or join two or more arrays into one. The only required argument is list or tuple of arrays. 💡For `np.concatenate()` to work, all the input array dimensions for the concatenation axis must match exactly, other wise you will get `ValueError` ### 1.1. 1D arrays Let’s get straight to work and define two arrays and then join them: ```python # first, import numpy import numpy as np # making two arrays to concatenate arr1 = np.arange(1,4) arr2 = np.arange(4,7) print("Arrays to concatenate:") print(arr1);print(arr2) print("After concatenation:") print(np.concatenate([arr1,arr2])) ``` ``` Arrays to concatenate: [1 2 3] [4 5 6] After concatenation: [1 2 3 4 5 6] ``` We can also join more than two arrays: ```python # we can concatenate more than two arrays arr3 = np.arange(7,10) np.concatenate([arr1,arr2,arr3]) ``` ``` array([1, 2, 3, 4, 5, 6, 7, 8, 9]) ``` ### 1.2. nD arrays The same constructor is used for joining two or more nD arrays By default, `np.concat` joins along the row wise (stack on top of each other). However, by providing kwarg `axis=1`, we can also concatenate along columns. ```python # creating 2D arrays to join ar2d1 = np.arange(1,7).reshape(2,3) ar2d2 = np.arange(7,13).reshape(2,3) print("2D arrays to concatenate:") print(ar2d1);print(ar2d2) # case1: joining along rows print("\nConcatenating along rows:") print(np.concatenate([ar2d1,ar2d2])) # case1: joining along columns print("\nConcatenating along columns:") print(np.concatenate([ar2d1,ar2d2], axis=1)) ``` ``` 2D arrays to concatenate: [[1 2 3] [4 5 6]] [[ 7 8 9] [10 11 12]] Concatenating along rows: [[ 1 2 3] [ 4 5 6] [ 7 8 9] [10 11 12]] Concatenating along columns: [[ 1 2 3 7 8 9] [ 4 5 6 10 11 12]] ``` ## 2. CONCATENATING ARRAYS OF MIXED DIMENSIONS To concatenate arrays of mixed dimensions we will use `np.vstack` and `np.hstack` functions ### 2.1. Vertical Stack ( np.vstack) > `np.vstack` concatenate along rows, stack vertically. 💡 To make `vstack` work, all input arrays should have same number of columns ```python # defining two 2D arrays v1 = np.arange(8).reshape(2,4) v2 = np.arange(20).reshape(5,4) print("We vstack these arrays:\n") print(f"Array 1: \n{v1}");print(f"Array 2: \n{v2}") # vstack print("\nAfter vStack-ed both arrays") print(np.vstack([v1,v2])) ``` ``` We will vstack these arrays: Array 1: [[0 1 2 3] [4 5 6 7]] Array 2: [[ 0 1 2 3] [ 4 5 6 7] [ 8 9 10 11] [12 13 14 15] [16 17 18 19]] After vStack-ed both arrays [[ 0 1 2 3] [ 4 5 6 7] [ 0 1 2 3] [ 4 5 6 7] [ 8 9 10 11] [12 13 14 15] [16 17 18 19]] ``` ### 2.1. Horizontal Stack ( np.hstack) > `np.hstack` concatenate along columns, stack one after the other. 💡To make `hstack` work, both arrays should have same number of rows ```python h1 = np.arange(10).reshape(5,2) h2 = np.arange(30).reshape(5,6) print("We will hstack these arrays:\n") print(f"Array 1: \n{h1}");print(f"Array 2: \n{h2}") print("\nAfter hStack-ed both arrays") print(np.hstack([h1,h2])) ``` ``` We will hstack these arrays: Array 1: [[0 1] [2 3] [4 5] [6 7] [8 9]] Array 2: [[ 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]] After hStack-ed both arrays [[ 0 1 0 1 2 3 4 5] [ 2 3 6 7 8 9 10 11] [ 4 5 12 13 14 15 16 17] [ 6 7 18 19 20 21 22 23] [ 8 9 24 25 26 27 28 29]] ``` > `np.dstack` is used for stacking 3 arrays ## 3. SPLITTING Opposite of concatenation is splitting and we will use `np.split`, `np.vsplit` and `np.hsplit` to split the arrays ### 3.1. np.split > `np.split` for N split points creates N+1, sub-arrays `np.split(x,y)` will split the array into three subarray. An example will make the concept clearer: ```python # array to splt ns = np.arange(1,10) print(f"Original array: {ns}") print("After split:") ns1,ns2,ns3 = np.split(ns, (3,6)) print(ns1);print(ns2);print(ns3) ``` ``` Original array: [1 2 3 4 5 6 7 8 9] After nsplit: [1 2 3] [4 5 6] [7 8 9] ``` ### 3.2… np.vsplit > `np.vsplit` splits along the vertical axis. You can either provide: * `integer` of equally shaped array (method 1 below), or, * by specifying the *row* `[integer]` at which the division should occur (method 2 below) ```python vs = np.arange(16).reshape(4,4) print(f"Array to vSplit: \n{vs}") # method 1, by specifying number of equally shaped arrays print("\nUsing Method 1") upper, lower = np.vsplit(vs, 2) print(f"Upper: \n{upper}\nLower: \n{lower}") # method 2, by specifying row, after which division should occur print("\nUsing Method 2") print(np.vsplit(vs, [2])) ``` ``` Array to vSplit: [[ 0 1 2 3] [ 4 5 6 7] [ 8 9 10 11] [12 13 14 15]] Using Method 1 Upper: [[0 1 2 3] [4 5 6 7]] Lower: [[ 8 9 10 11] [12 13 14 15]] Using Method 2 [array([[0, 1, 2, 3], [4, 5, 6, 7]]), array([[ 8, 9, 10, 11], [12, 13, 14, 15]])] ``` ### 3.3. np.hsplit > `np.hsplit` splits along the horizontal axis. You can either provide: * number `integer` of equally shaped array, or, * by specifying the *column* `[integer]` at which the division should occur ```python hs = np.arange(16).reshape(2,8) print(f"Array to hSplit: \n{hs}") # method 1, by specifying number of equally shaped arrays print("\nUsing Method 1") right, left = np.hsplit(hs, 2) print(f"Right: \n{right}\nLeft: \n{left}") # method 2, by specifying columns, after which division should occur print("\nUsing Method 2") print(np.hsplit(hs, [4])) ``` ``` Array to hSplit: [[ 0 1 2 3 4 5 6 7] [ 8 9 10 11 12 13 14 15]] Using Method 1 Right: [[ 0 1 2 3] [ 8 9 10 11]] Left: [[ 4 5 6 7] [12 13 14 15]] Using Method 2 [array([[ 0, 1, 2, 3], [ 8, 9, 10, 11]]), array([[ 4, 5, 6, 7], [12, 13, 14, 15]])] ``` --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://codingnotes.gitbook.io/coding_notes/coding/numpy/concatenation-and-splitting.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.