{"id":39260,"date":"2023-01-10T17:05:12","date_gmt":"2023-01-10T17:05:12","guid":{"rendered":"https:\/\/www.askpython.com\/?p=39260"},"modified":"2023-02-16T19:56:25","modified_gmt":"2023-02-16T19:56:25","slug":"numpy-ldexp","status":"publish","type":"post","link":"https:\/\/www.askpython.com\/python-modules\/numpy\/numpy-ldexp","title":{"rendered":"NumPy ldexp: A Complete Guide"},"content":{"rendered":"\n

Hello! In this tutorial, we are going learn about the ldexp <\/strong>method present in the NumPy module in Python. In our previous tutorials, we have learnt about the NumPy functions exp()<\/a> and exp2()<\/a>. <\/p>\n\n\n\n

We know that, the exp(x)<\/code> function returns e^x<\/code>, where e<\/em> is Euler’s constant having the value of approximately 2.718281 and the exp2(x)<\/code> function calculates the value 2^x<\/code>. <\/p>\n\n\n\n

Now, let us learn about the ldexp()<\/code> function in NumPy.<\/p>\n\n\n\n

\n\n\n\n

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

The ldexp<\/code> function in NumPy takes two parameters x1 <\/em>and x2<\/em> and returns x1*2**x2<\/strong><\/em>, element-wise. Mathematically, we can represent it as follows:<\/p>\n\n\n\n

\"Ldexp
Ldexp Equation<\/figcaption><\/figure>\n\n\n\n

Note that, in Python, the **<\/strong> operator has more precedence than the *<\/strong> operator. Hence, first the value 2**x2<\/em><\/strong> is calculated and then it is multiplied by x1<\/em><\/strong> to compute the final result.<\/p>\n\n\n\n

\n\n\n\n

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

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

\nnumpy.ldexp(x1, x2, \/, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])\n<\/pre><\/div>\n\n\n
Parameter<\/strong><\/td>Description<\/strong><\/td>Required\/Optional<\/strong><\/td><\/tr>
x1 (array_like)<\/em><\/td>The array of multipliers.<\/td>Required<\/td><\/tr>
x2 (array_like)<\/em><\/td>The array of exponents. <\/td>Required<\/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>
where

<\/td>		Takes an array-like object. At locations where it is True, the out<\/code> array will be set to the ufunc<\/code> result. Elsewhere, the out<\/code> array will retain its original value.<\/td>Optional<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
The shape of x1<\/em> and x2<\/em> should be equal.<\/p>\n\n\n\n
Returns:<\/strong> <\/p>\n\n\n\n
An n-dimensional array storing the result of x1***x2<\/strong><\/em>. This is scalar if both x1<\/strong><\/em> and x2<\/em><\/strong> are scalars.<\/p>\n\n\n\n
Note that<\/strong>, complex data types are not supported and if provided, will raise a TypeError<\/em>.<\/p>\n\n\n\n
\n\n\n\n
Examples of NumPy ldexp<\/h2>\n\n\n\n
Now let us get right into the examples and understand how the function works.<\/p>\n\n\n\n
1. When both the inputs are scalars<\/h3>\n\n\n
\nimport numpy as np\n\n# case 1\nx1_1 = 2\nx2_1 = 3\nans_1 = np.ldexp(x1_1, x2_1)\nprint("Case 1:\\nx1 = ", x1_1, ", x2 = ", x2_1, ", Result = ", ans_1)\n\n# case 2\nx1_2 = -4\nx2_2 = 5\nans_2 = np.ldexp(x1_2, x2_2)\nprint("Case 2:\\nx1 = ", x1_2, ", x2 = ", x2_2, ", Result = ", ans_2)\n\n# case 3\nx1_3 = 6.8\nx2_3 = 2\nans_3 = np.ldexp(x1_3, x2_3)\nprint("Case 3:\\nx1 = ", x1_3, ", x2 = ", x2_3, ", Result = ", ans_3)\n<\/pre><\/div>\n\n\n
Output:<\/strong><\/p>\n\n\n
\nCase 1:\nx1 =  2 , x2 =  3 , Result =  16.0\nCase 2:\nx1 =  -4 , x2 =  5 , Result =  -128.0\nCase 3:\nx1 =  6.8 , x2 =  2 , Result =  27.2\n<\/pre><\/div>\n\n\n
The outputs are computed as follows:<\/p>\n\n\n\n
Case 1:<\/strong><\/p>\n\n\n
\n2**x2  = 2**3 = 8\nResult = x1*8 = 2*8 = 16\n<\/pre><\/div>\n\n\n
Case 2:<\/strong><\/p>\n\n\n
\n2**x2  = 2**5 = 32\nResult = x1*32 = (-4)*32 = -128\n<\/pre><\/div>\n\n\n
Case 3:<\/strong><\/p>\n\n\n
\n2**x2  = 2**2 = 4\nResult = x1*4 = 6.8*4 = 27.2\n<\/pre><\/div>\n\n\n
\n\n\n\n
2. When both the inputs are arrays<\/h3>\n\n\n
\nimport numpy as np\n\nlist_1 = [-5, 2, 4, 6]\nlist_2 = [4, 3, -2, 1]\n\nans = np.ldexp(list_1, list_2)\nprint(ans)\n<\/pre><\/div>\n\n\n
Output:<\/strong><\/p>\n\n\n
\n[-80.  16.   1.  12.]\n<\/pre><\/div>\n\n\n
The computation of the output is as follows:<\/p>\n\n\n
\nans[0] = -5*2**4 = -5*16 = -80\nans[1] = 2*2**3 = 2*8 = 16\nans[2] = 4*2**-2 = 4*0.25 = 1\nans[3] = 6*2**1 = 6*2 = 12\n<\/pre><\/div>\n\n\n
\n\n\n\n
3. When one input is an array and the other is a  scalar<\/h3>\n\n\n
\nimport numpy as np\n\nl = [8, 12, 3, 10]\nx = 3\n\nans = np.ldexp(l, x)\nprint(ans)\n<\/pre><\/div>\n\n\n
Output:<\/strong><\/p>\n\n\n
\n[64. 96. 24. 80.]\n<\/pre><\/div>\n\n\n
Here, since the second argument x<\/em><\/strong> is a scalar, for each value in the answer, the element in the list is multiplied by 2**x<\/em><\/strong> i.e. 2**3 = 8.<\/em><\/strong><\/p>\n\n\n
\nans[0] = 8*2**3 = 8*8 = 64\nans[1] = 12*2**3 = 12*8 = 96\nans[2] = 3*2**3 = 3*8 = 24\nans[3] = 10*2**3 = 10*8 = 80\n<\/pre><\/div>\n\n\n
\n\n\n\n
Conclusion<\/h2>\n\n\n\n
That\u2019s all! In this tutorial, we learned about the Numpy ldexp<\/strong> method and practiced different types of examples using the same. If you want to learn more about NumPy, feel free to go through our NumPy tutorials<\/a>.<\/p>\n\n\n\n
\n\n\n\n
Reference<\/h2>\n\n\n\n