{"id":38038,"date":"2022-12-16T15:44:11","date_gmt":"2022-12-16T15:44:11","guid":{"rendered":"https:\/\/www.askpython.com\/?p=38038"},"modified":"2023-02-16T19:56:27","modified_gmt":"2023-02-16T19:56:27","slug":"numpy-nanmin","status":"publish","type":"post","link":"https:\/\/www.askpython.com\/python-modules\/numpy\/numpy-nanmin","title":{"rendered":"NumPy nanmin \u2013 Minimum of an array along an axis ignoring any NaNs"},"content":{"rendered":"\n

Hello and welcome to this tutorial on\u00a0Numpy nanmin<\/strong>. In this tutorial, we will be learning about the NumPy\u00a0nanmin()<\/code>\u00a0method and also seeing a lot of examples regarding the same. So let us begin!

Also read:\u00a0NumPy nanmax \u2013 Maximum of an array along an axis ignoring any NaNs<\/a><\/em><\/strong><\/p>\n\n\n\n

\n\n\n\n

What is NumPy nanmin?<\/h2>\n\n\n\n

In Python, NaN<\/strong> denotes Not a Number<\/strong>. If we have an array that contains some NaN values and want to find the minimum value in it, we can use the nanmin()<\/code> method from NumPy.<\/p>\n\n\n\n

The\u00a0nanmin()<\/code>\u00a0method in NumPy is a function that returns the minimum of the array elements calculated by ignoring the NaN values in the array. It can be the minimum of all the array elements, the minimum of the array elements along the rows or the minimum of the array elements along the columns. <\/p>\n\n\n\n

We will see the examples for each of these in the upcoming section of this tutorial.<\/p>\n\n\n\n

\n\n\n\n

Syntax of NumPy nanmin<\/h2>\n\n\n\n

Let us have a look at the syntax of the nanmin()<\/code> function.<\/p>\n\n\n

\nnumpy.nanmin(a, axis=None, out=None, keepdims=<no value>, initial=<no value>, where=<no value>)\n<\/pre><\/div>\n\n\n
Parameter<\/strong><\/td>Description<\/strong><\/td>Required\/Optional<\/strong><\/td><\/tr>
a (array_like)<\/td>Input data.<\/td>Required<\/td><\/tr>
axis<\/td>Axis along which the minimum of the array is to be calculated.
It can be axis=0 or axis=1 or axis=None which implies that the minimum of the entire array is to be returned.<\/td>		Optional<\/td><\/tr>
out<\/td>An alternative output array in which to place the result. It must have the same shape as the expected output.<\/td>Optional<\/td><\/tr>
keepdims (bool)<\/td>If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the input array.<\/td>Optional<\/td><\/tr>
initial<\/td>The maximum value of an output element.<\/td>Optional<\/td><\/tr>
where<\/td>Elements to compare for finding the minimum.<\/td>Optional<\/td><\/tr><\/tbody><\/table>
NumPy nanmin parameters<\/figcaption><\/figure>\n\n\n\n
Returns:<\/strong>
An array containing the minimum of the array along the specified axis, ignoring all the NaNs.<\/p>\n\n\n\n
\n\n\n\n
Examples of NumPy nanmin<\/h2>\n\n\n\n
Let\u2019s get right into the different examples of using numpy.nanmin()<\/strong> function.<\/p>\n\n\n\n
NumPy nanmin of a 1-d array<\/h3>\n\n\n
\nimport numpy as np\n\narr = [np.nan, 54, 1, 3, 44]\n# using the nanmin function to calculate the maximum\nans = np.nanmin(arr)\n\nprint("arr =", arr)\nprint("Result =", ans)\n<\/pre><\/div>\n\n\n
Output:<\/strong><\/p>\n\n\n
\narr = [nan, 54, 1, 3, 44]\nResult = 1.0\n<\/pre><\/div>\n\n\n
Ignoring the NaN value, the minimum value among 54, 1, 3 and 44 is 1, hence it is returned.<\/p>\n\n\n\n
\n\n\n\n
NumPy nanmin of a 2-d array<\/h3>\n\n\n
\nimport numpy as np\n\narr = [[30, -9], [8, np.nan]]\n# using the nanmin function to calculate the maximum\nans = np.nanmin(arr)\n\nprint("arr =", arr)\nprint("Result =", ans)\n<\/pre><\/div>\n\n\n
Output:<\/strong><\/p>\n\n\n
\narr = [[30, -9], [8, nan]]\nResult = -9.0\n<\/pre><\/div>\n\n\n
Similar to the previous example, the minimum of 30, -9, and 8 is 8.<\/p>\n\n\n\n
\n\n\n\n
NumPy nanmin along an axis of the array<\/h3>\n\n\n\n
axis = 0<\/strong><\/p>\n\n\n
\nimport numpy as np\n\narr = [[16, 4], [np.nan, 1]]\n# using the nanmin function to calculate the maximum\nans = np.nanmin(arr, axis=0)\n\nprint("arr =", arr)\nprint("Result =", ans)\n<\/pre><\/div>\n\n\n
Output:<\/strong><\/p>\n\n\n
\narr = [[16, 4], [nan, 1]]\nResult = [16.  1.]\n<\/pre><\/div>\n\n\n
Here, the values in each row for a specific column are compared to find the minimum element.<\/p>\n\n\n
\nans[0] = min(arr[0][0], arr[1][0]) = min(16, np.nan) = 16 (ignoring NaN)\nans[1] = min(arr[0][1], arr[1][1]) = min(4, 1) = 1\n<\/pre><\/div>\n\n\n
axis = 1<\/strong><\/p>\n\n\n
\nimport numpy as np\n\narr = [[16, 4], [np.nan, 1]]\n# using the nanmin function to calculate the maximum\nans = np.nanmin(arr, axis=1)\n\nprint("arr =", arr)\nprint("Result =", ans)\n<\/pre><\/div>\n\n\n
Output:<\/strong><\/p>\n\n\n
\narr = [[16, 4], [nan, 1]]\nResult = [4. 1.]\n<\/pre><\/div>\n\n\n
When axis=1<\/em>, the elements in each row are compared over all the columns to find the minimum.<\/p>\n\n\n
\nans[0] = min(arr[0][0], arr[0][1]) = min(16, 4) = 4\nans[1] = min(arr[1][0], arr[1][1]) = min(np.nan, 1) = 1 (ignoring NaN)\n<\/pre><\/div>\n\n\n
\n\n\n\n
NumPy nanmin of an array containing infinity<\/h3>\n\n\n\n
Let us now see how the numpy.nanmin()<\/code> method handles infinity along with NaNs in the array.<\/p>\n\n\n
\nimport numpy as np\n \n# array containing +infinity\na = np.array([16, 3, np.nan, 7, np.inf])\n# array containing -infinity\nb = np.array([16, 3, np.nan, 7, np.NINF])\n# array containing +infinity and -infinity\nc = np.array([16, 3, np.nan, np.NINF, 7, np.inf])\n \nmin_a = np.nanmin(a)\nmin_b = np.nanmin(b)\nmin_c = np.nanmin(c)\n \nprint("a =", a)\nprint("Minimum of the array a =", min_a)\nprint("\\nb =", b)\nprint("Minimum of the array b =", min_b)\nprint("\\nc =", c)\nprint("Minimum of the array c =", min_c)\n<\/pre><\/div>\n\n\n
Output:<\/strong><\/p>\n\n\n
\na = [16.  3. nan  7. inf]\nMinimum of the array a = 3.0\n\nb = [ 16.   3.  nan   7. -inf]\nMinimum of the array b = -inf\n\nc = [ 16.   3.  nan -inf   7.  inf]\nMinimum of the array c = -inf\n<\/pre><\/div>\n\n\n
In the above code, NINF<\/strong> denotes -infinity <\/strong>and inf <\/strong>denotes infinity<\/strong>. Note that,<\/p>\n\n\n\n