{
"cells": [
{
"cell_type": "markdown",
"id": "f20b26d7",
"metadata": {},
"source": [
"--- \n",
" ![](\"https://upload.wikimedia.org/wikipedia/en/c/c8/University_of_the_Punjab_logo.png\")
\n",
"\n",
"Department of Data Science
\n",
"Course: Tools and Techniques for Data Science
\n",
"\n",
"---\n",
"Instructor: Muhammad Arif Butt, Ph.D.
"
]
},
{
"cell_type": "markdown",
"id": "3f7ee3b8",
"metadata": {},
"source": [
"Lecture 3.6 (NumPy-06)
"
]
},
{
"cell_type": "markdown",
"id": "64ba925a",
"metadata": {},
"source": [
"![\"Open](\"https://colab.research.google.com/assets/colab-badge.svg\")
"
]
},
{
"cell_type": "markdown",
"id": "fa6dcb19",
"metadata": {},
"source": [
"# _Manipulating Array Elements.ipynb_"
]
},
{
"cell_type": "markdown",
"id": "c97f0f67",
"metadata": {},
"source": [
"![](\"images/numpy123D.png\")
\n",
"\n",
"# Learning agenda of this notebook\n",
"\n",
"1. Updating existing values of NumPy array elements\n",
"2. Append new elements to a NumPy array using np.append()\n",
"3. Insert new elements in a NumPy array using np.insert()\n",
"4. Delete elements of a NumPy array using np.delete()\n",
"5. Alias vs Shallow Copy vs Deep Copy"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "eaf623d0",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "5f9a8b93",
"metadata": {},
"outputs": [],
"source": [
"# To install this library in Jupyter notebook\n",
"#import sys\n",
"#!{sys.executable} -m pip install numpy"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "6fa14827",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"('1.19.5',\n",
" ['/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/numpy'])"
]
},
"execution_count": 1,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import numpy as np\n",
"np.__version__ , np.__path__"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "a55cc9b4",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "9800e96a",
"metadata": {},
"source": [
"## 1. Updating Existing Values of Numpy Array Elements"
]
},
{
"cell_type": "markdown",
"id": "2c75b1aa",
"metadata": {},
"source": [
"### a. 1-D Arrays"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "c22d47e9",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Original Array \n",
" [21 64 42 91 22 79 40 84 91 92]\n",
"Updated Array \n",
" [ 21 64 333 91 22 79 40 84 91 777]\n"
]
}
],
"source": [
"arr = np.random.randint(low = 1, high = 100, size = 10)\n",
"print(\"Original Array \\n\", arr)\n",
"arr[2] = 333\n",
"arr[-1] = 777\n",
"print(\"Updated Array \\n\", arr)"
]
},
{
"cell_type": "markdown",
"id": "2bd1f603",
"metadata": {},
"source": [
"### b. 2-D Arrays"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "3e3b0f21",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Original Array \n",
" [[7 4 7 1]\n",
" [3 9 2 3]\n",
" [7 4 8 5]\n",
" [4 5 5 9]]\n",
"Updated Array \n",
" [[ 7 77 7 1]\n",
" [ 3 9 66 3]\n",
" [ 7 4 8 22]\n",
" [ 4 5 5 9]]\n"
]
}
],
"source": [
"# Creating 2-D array of size 4x4 of int type b/w interval 1 to 9\n",
"arr = np.random.randint(low = 1, high = 10, size = (4, 4))\n",
"print(\"Original Array \\n\", arr)\n",
"arr[0][1] = 77\n",
"arr[1][2] = 66\n",
"arr[2][-1] = 22\n",
"print(\"Updated Array \\n\", arr)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "ec4398ea",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "73186f77",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "a22f0d64",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "edf452b4",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "c7b21ead",
"metadata": {},
"source": [
"## 2. Append New Elements to Numpy Arrays\n",
"- The `np.append()` method allows us to insert new values at the end of a NumPy array.\n",
"- The method always returns a copy of the existing numpy array with the values appended to the given axis.\n",
"```\n",
"np.append(arr, values, axis=None)\n",
"```\n",
"- Where,\n",
" - `arr` is the array in which we want to append\n",
" - `values` must be of the correct shape (the same shape as `arr` excluding `axis`)\n",
" - If `axis` is not specified, both `arr` and `values` are flattened before use.\n",
" - If `axis` is specified, then `values` must be of the correct shape (the same shape as `arr` excluding `axis`)\n",
"- The original array remains as such, as it does not occur in-place."
]
},
{
"cell_type": "markdown",
"id": "c9b7fa54",
"metadata": {},
"source": [
"### a. Appending Elements in 1-D Arrays"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "22b89811",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = [27 43 66 95 39 38 42 7 25 37]\n"
]
}
],
"source": [
"arr1 = np.random.randint(low = 1, high = 100, size = 10)\n",
"print(\"arr1 = \", arr1)"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "b966077f",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"After append:\n",
"arr1 = [27 43 66 95 39 38 42 7 25 37]\n",
"arr2 = [ 27 43 66 95 39 38 42 7 25 37 101 202 303]\n"
]
}
],
"source": [
"# You can add a scalar value or a list of values at the end of a 1-D array\n",
"arr2 = np.append(arr1, [101, 202,303])\n",
"print(\"After append:\")\n",
"print(\"arr1 = \", arr1)\n",
"print(\"arr2 = \", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "f809838c",
"metadata": {},
"outputs": [],
"source": [
"print(id(arr1))\n",
"print(id(arr2))"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "f29ee507",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "c42b712a",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "196785fd",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "b71d3c60",
"metadata": {},
"source": [
"### b. Appending Elements in 2-D Arrays"
]
},
{
"cell_type": "markdown",
"id": "94d9cf4a",
"metadata": {},
"source": [
"**Example:** In case of 2-D Arrays if `axis` is not mentioned both `arr` and `values` are flattened before use"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "9fdcd017",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = \n",
" [[3 3 1]\n",
" [7 7 3]\n",
" [6 8 7]]\n"
]
}
],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = (3,3))\n",
"print(\"arr1 = \\n\", arr1)"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "dfdb6a95",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"After append:\n",
"arr1 = \n",
" [[3 3 1]\n",
" [7 7 3]\n",
" [6 8 7]]\n",
"arr2 = [ 3 3 1 7 7 3 6 8 7 101 202 303 404 505]\n"
]
}
],
"source": [
"# If the axis is not mentioned, values can be of any shape and both `arr` and `values` are flattened before use.\n",
"arr2 = np.append(arr1, [101, 202,303, 404, 505])\n",
"print(\"After append:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "302c7f71",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "137dc887",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "c5de2401",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "8c44d2ef",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "130e285f",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "28824ef2",
"metadata": {},
"source": [
"**Example:** Appending a Row to a 2-D array (`axis=0`)"
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "469af8c2",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = \n",
" [[2 7 2]\n",
" [8 7 3]\n",
" [4 1 9]\n",
" [9 8 1]]\n",
"shape: (4, 3)\n"
]
}
],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = (4,3))\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"shape: \", arr1.shape)"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "43aea3c8",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"After append:\n",
"arr1 = \n",
" [[2 7 2]\n",
" [8 7 3]\n",
" [4 1 9]\n",
" [9 8 1]]\n",
"arr2 = \n",
" [[ 2 7 2]\n",
" [ 8 7 3]\n",
" [ 4 1 9]\n",
" [ 9 8 1]\n",
" [101 202 303]]\n"
]
}
],
"source": [
"# For appending at axis 0, the values argument must the same shape as `arr` excluding `axis`\n",
"# so the values should be a row vector, and in this case of shape (1,3), having 1 row and 3 columns\n",
"arr2 = np.append(arr1, [[101, 202,303]], axis=0)\n",
"print(\"After append:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \\n\", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "d8b37d3f",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "b385e069",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "8d04630e",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "d820ca37",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "154a8a0a",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "1e3ef38e",
"metadata": {},
"source": [
"**Example:** Appending a Column to a 2-D array (`axis=1`)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "70c4b2dc",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = \n",
" [[4 5 6]\n",
" [8 3 5]\n",
" [4 6 9]\n",
" [4 2 7]]\n",
"shape: (4, 3)\n"
]
}
],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = (4,3))\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"shape: \", arr1.shape)"
]
},
{
"cell_type": "code",
"execution_count": 11,
"id": "4d32417b",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"After append:\n",
"arr1 = \n",
" [[4 5 6]\n",
" [8 3 5]\n",
" [4 6 9]\n",
" [4 2 7]]\n",
"arr2 = \n",
" [[ 4 5 6 101]\n",
" [ 8 3 5 202]\n",
" [ 4 6 9 303]\n",
" [ 4 2 7 404]]\n"
]
}
],
"source": [
"# For appending at axis 1, the values argument must the same shape as `arr` excluding `axis`\n",
"# so the values should be a column vector of shape (4,1), having 4 rows and 1 column\n",
"arr2 = np.append(arr1, [[101], [202], [303], [404]], axis=1)\n",
"print(\"After append:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \\n\", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "277823a5",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "a969f309",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "eaadbd59",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "66964303",
"metadata": {},
"source": [
"## 3. Inserting New Elements in Numpy Arrays\n",
"- The `np.insert()` method allows us to insert new values along the given axis before the given index.\n",
"- The method always returns a copy of the existing numpy array with the values inserted to the given axis.\n",
"```\n",
"np.insert(arr, index, values, axis=None)\n",
"```\n",
"- Where,\n",
" - `arr` is the array in which we want to insert\n",
" - `index` is the index before which we want to insert\n",
" - `values` [array_like] values to be added in the `arr`\n",
" - If `axis` is not specified, both `arr` and `values` are flattened before use.\n",
" - If `axis` is zero, a row is inserted (For 2-D arrays)\n",
" - If `axis` is one, a column is inserted (For 2-D arrays)\n",
"- The original array remains as such, as it does not occur in-place."
]
},
{
"cell_type": "markdown",
"id": "962d7c69",
"metadata": {},
"source": [
"### a. Inserting Elements in 1-D Arrays"
]
},
{
"cell_type": "code",
"execution_count": 12,
"id": "668f2e4a",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = [15 58 34 72 77]\n"
]
}
],
"source": [
"arr1 = np.random.randint(low = 1, high = 100, size = 5)\n",
"print(\"arr1 = \", arr1)"
]
},
{
"cell_type": "code",
"execution_count": 13,
"id": "a6c59fdd",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"After insert:\n",
"arr1 = [15 58 34 72 77]\n",
"arr2 = [15 58 34 55 66 77 72 77]\n"
]
}
],
"source": [
"# You can insert a scalar value or a list of values in between array elements before the mentioned index\n",
"arr2 = np.insert(arr1, 3, [55, 66,77])\n",
"print(\"After insert:\")\n",
"print(\"arr1 = \", arr1)\n",
"print(\"arr2 = \", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "a513d985",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "5edc791f",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "3bf0a3ab",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "0e468e72",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "06a5cb18",
"metadata": {},
"source": [
"### b. Inserting Elements in 2-D Arrays"
]
},
{
"cell_type": "markdown",
"id": "de9ed861",
"metadata": {},
"source": [
"**Example:** In case of 2-D array, if `axis` is not mentioned the array is flattened first"
]
},
{
"cell_type": "code",
"execution_count": 14,
"id": "3284a269",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = \n",
" [[5 5 1 9]\n",
" [8 3 1 1]\n",
" [2 7 9 2]]\n"
]
}
],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = (3,4))\n",
"print(\"arr1 = \\n\", arr1)"
]
},
{
"cell_type": "code",
"execution_count": 15,
"id": "cfad3dcf",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"After insert:\n",
"arr1 = \n",
" [[5 5 1 9]\n",
" [8 3 1 1]\n",
" [2 7 9 2]]\n",
"arr2 = [ 5 5 1 9 55 8 3 1 1 2 7 9 2]\n"
]
}
],
"source": [
"# Inserting a single value\n",
"arr2 = np.insert(arr1, 4, 55)\n",
"print(\"After insert:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "40317d8a",
"metadata": {},
"outputs": [],
"source": [
"# Inserting a multiple values\n",
"arr2 = np.insert(arr1, 4, [55, 66])\n",
"print(\"After insert:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "ade890b3",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "8ef7679e",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "556b2299",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "816fd274",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "f2a9adba",
"metadata": {},
"source": [
"**Example:** If axis=0, value(s) are added as a row before mentioned index"
]
},
{
"cell_type": "code",
"execution_count": 16,
"id": "fc4f9a6b",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = \n",
" [[8 9 3 7]\n",
" [9 8 7 7]\n",
" [8 6 1 9]]\n"
]
}
],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = (3,4))\n",
"print(\"arr1 = \\n\", arr1)"
]
},
{
"cell_type": "code",
"execution_count": 17,
"id": "a11a09b7",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"After insert:\n",
"arr1 = \n",
" [[8 9 3 7]\n",
" [9 8 7 7]\n",
" [8 6 1 9]]\n",
"arr2 = \n",
" [[ 8 9 3 7]\n",
" [ 9 8 7 7]\n",
" [55 55 55 55]\n",
" [ 8 6 1 9]]\n"
]
}
],
"source": [
"# For axis = 0, note how the scalar value is replicated before insertion\n",
"arr2 = np.insert(arr1, 2, 55, axis=0)\n",
"print(\"After insert:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \\n\", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "49fbdaad",
"metadata": {},
"outputs": [],
"source": [
"# For axis=0, note the size of values has to be 4 in this case (equal to number of columns)\n",
"arr2 = np.insert(arr1, 2, [55, 66, 77, 88], axis=0)\n",
"print(\"After insert:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \\n\", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "bfca2c79",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "8f584165",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "0ea4dcf0",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "87670040",
"metadata": {},
"source": [
"**Example:** If axis=1, value(s) are added as a column at mentioned index"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "a92e0c42",
"metadata": {},
"outputs": [],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = (3,4))\n",
"print(\"arr1 = \\n\", arr1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "e6d02c09",
"metadata": {},
"outputs": [],
"source": [
"# For axis = 1, note how the scalar value is replicated before insertion\n",
"arr2 = np.insert(arr1, 2, 55, axis=1)\n",
"print(\"After insert:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \\n\", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "08c536be",
"metadata": {},
"outputs": [],
"source": [
"# For axis=1, note the size of values has to be 3 in this case (equal to number of rows)\n",
"arr2 = np.insert(arr1, 2, [55, 66, 77], axis=1)\n",
"print(\"After insert:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \\n\", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "a9c796a3",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "835da82c",
"metadata": {},
"source": [
"## 4. Deleting Elements of Numpy Arrays\n",
"- The `np.delete()` method allows us to delete value(s) from an array at the given index\n",
"- This function always returns a copy of the existing numpy array with the values deleted from the given axis.\n",
"- If axis is not specified, values can be of any shape and will be flattened before use\n",
"```\n",
"np.delete(arr, index, axis=None)\n",
"```\n",
"- The original array remains as such, as it does not occur in-place."
]
},
{
"cell_type": "markdown",
"id": "8179e2e5",
"metadata": {},
"source": [
"### a. Deleting Elements from a 1-D Arrays"
]
},
{
"cell_type": "markdown",
"id": "0e6c6747",
"metadata": {},
"source": [
"**Example:**"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "fe38b972",
"metadata": {},
"outputs": [],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = 5)\n",
"print(\"arr1 = \", arr1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "e988364b",
"metadata": {},
"outputs": [],
"source": [
"# You can delete a scalar value from a specific index\n",
"arr2 = np.delete(arr1, 3)\n",
"print(\"After delete:\")\n",
"print(\"arr1 = \", arr1)\n",
"print(\"arr2 = \", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "d8b82c88",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "c3ce1cc1",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "123f4850",
"metadata": {},
"source": [
"**Example:**"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "e1911555",
"metadata": {},
"outputs": [],
"source": [
"arr1 = np.random.randint(low = 1, high = 100, size = 10)\n",
"print(\"arr1 = \", arr1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "73ce969a",
"metadata": {},
"outputs": [],
"source": [
"# You can delete a list of values in between array elements from specific indices\n",
"arr2 = np.delete(arr1, [2,5])\n",
"print(\"After delete:\")\n",
"print(\"arr1 = \", arr1)\n",
"print(\"arr2 = \", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "10436aac",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "20dd4fec",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "d8cf1d87",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "569a9b74",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "2f5fefe0",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "9fd1c3ac",
"metadata": {},
"source": [
"### b. Deleting Elements from a 2-D Arrays"
]
},
{
"cell_type": "markdown",
"id": "6bc492dc",
"metadata": {},
"source": [
"**Example:** Delete a specific element from a 2-D array, don't mention the axis. The resulting array is flattened before use"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "3b859729",
"metadata": {},
"outputs": [],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = (3,3))\n",
"print(\"arr1 = \\n\", arr1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "fe24666f",
"metadata": {},
"outputs": [],
"source": [
"arr2 = np.delete(arr1, 5)\n",
"print(\"After delete:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "2dcac4ef",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "a179b1ec",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "2dad6e7d",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "fa78d4b7",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "25e85c4f",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "572970be",
"metadata": {},
"source": [
"**Example:** Delete a specific row from an existing 2-D array"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "9c787db6",
"metadata": {},
"outputs": [],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = (4,4))\n",
"print(\"arr1 = \\n\", arr1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "9aa44191",
"metadata": {},
"outputs": [],
"source": [
"arr2 = np.delete(arr1, 2, axis=0)\n",
"print(\"After delete:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \\n\", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "46b11add",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "467e9c34",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "c59461c9",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "4c2a31aa",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "12e50b96",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "a67df0fe",
"metadata": {},
"source": [
"**Example:** Delete a specific column from an existing 2-D array"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "3e08589e",
"metadata": {},
"outputs": [],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = (4,4))\n",
"print(\"arr1 = \\n\", arr1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "e6da91c2",
"metadata": {},
"outputs": [],
"source": [
"arr2 = np.delete(arr1, 2, axis=1)\n",
"print(\"After delete:\")\n",
"print(\"arr1 = \\n\", arr1)\n",
"print(\"arr2 = \\n\", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "977612c1",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "4d25f455",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "770bac4c",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "7f00bded",
"metadata": {},
"source": [
"## 5. Assigning vs Coping NumPy Arrays"
]
},
{
"cell_type": "markdown",
"id": "8e0e6a49",
"metadata": {},
"source": [
"### a. Assigning two NumPy Arrays (Create an alias)"
]
},
{
"cell_type": "code",
"execution_count": 18,
"id": "382c3f2c",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = [8 6 3 4 9 3 4 3 2 9]\n",
"arr2 = [8 6 3 4 9 3 4 3 2 9]\n",
"140657482813808\n",
"140657482813808\n"
]
}
],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = 10)\n",
"\n",
"# Creating a copy using assignment operator, both variables point at the same array\n",
"arr2 = arr1\n",
"\n",
"print(\"arr1 = \", arr1)\n",
"print(\"arr2 = \", arr2)\n",
"print(id(arr1))\n",
"print(id(arr2))"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "2ec4f294",
"metadata": {},
"outputs": [],
"source": [
"# Change value in arr1 will also occur in arr2\n",
"arr2[2] = 55\n",
"print(\"arr1 = \", arr1)\n",
"print(\"arr2 = \", arr2)\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "b8fd525a",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "d5dffdcd",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "aa2d253d",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "d6aa14c7",
"metadata": {},
"source": [
"### b. View/Shallow Copy\n",
"Arrays that share some data. The view method creates an object looking at the same data. Slicing an array returns a view of that array."
]
},
{
"cell_type": "code",
"execution_count": 19,
"id": "670adc44",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = [6 9 5 9 6 5 4 8 3 3]\n",
"arr2 = [6 9 5 9 6 5 4 8 3 3]\n",
"140657482814368\n",
"140657482813312\n"
]
}
],
"source": [
"import numpy as np\n",
"arr1 = np.random.randint(low = 1, high = 10, size = 10)\n",
"\n",
"# Creating a shallow copy (view) using slice operator\n",
"arr2 = arr1[:]\n",
"\n",
"print(\"arr1 = \", arr1)\n",
"print(\"arr2 = \", arr2)\n",
"print(id(arr1))\n",
"print(id(arr2))"
]
},
{
"cell_type": "code",
"execution_count": 20,
"id": "ede79da2",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = [ 6 9 55 9 6 5 4 8 3 3]\n",
"arr2 = [ 6 9 55 9 6 5 4 8 3 3]\n"
]
}
],
"source": [
"# Change value in arr1 will occur in arr2\n",
"arr2[2] = 55\n",
"\n",
"print(\"arr1 = \", arr1)\n",
"print(\"arr2 = \", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "2c5bf0ed",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "3371fba8",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "341a0304",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "markdown",
"id": "3f007f4f",
"metadata": {},
"source": [
"### c. Deep Copy"
]
},
{
"cell_type": "code",
"execution_count": 21,
"id": "1abe5e75",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = [1 5 9 4 1 5 9 8 6 6]\n",
"arr2 = [1 5 9 4 1 5 9 8 6 6]\n",
"140657482815488\n",
"140657482769472\n"
]
}
],
"source": [
"arr1 = np.random.randint(low = 1, high = 10, size = 10)\n",
"\n",
"# Create a Deep copy using copy() method, which will create a new copy of the array\n",
"arr2 = arr1.copy()\n",
"print(\"arr1 = \", arr1)\n",
"print(\"arr2 = \", arr2)\n",
"print(id(arr1))\n",
"print(id(arr2))"
]
},
{
"cell_type": "code",
"execution_count": 22,
"id": "5b11e005",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"arr1 = [1 5 9 4 1 5 9 8 6 6]\n",
"arr2 = [ 1 5 55 4 1 5 9 8 6 6]\n"
]
}
],
"source": [
"# Change value in array 1 will NOT occur in array 2\n",
"arr2[2] = 55\n",
"\n",
"print(\"arr1 = \", arr1)\n",
"print(\"arr2 = \", arr2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "5ec509b4",
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"id": "06a4ad16",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.9.7"
}
},
"nbformat": 4,
"nbformat_minor": 5
}