NumPy | Recipes reference

To access a value in a Numpy 2D array, use the [] syntax.

To access particular columns of a 2D Numpy array, use the [] syntax.

In Numpy, we use the [] syntax to access particular rows of a 2D Numpy array.

To calculate the determinant of a matrix in Python, use NumPy's linalg.det(~) method.

To check the allowed values for a NumPy integer type use the iinfo(~) method. For float types use the finfo(~) method.

To check if a NumPy array is a view or not use the base property. To check if a NumPy array is a copy, we can use np.shares_memory(~) to check whether the two objects share memory or not.

To check the version of your NumPy use print(np.__version__).

To check whether a NumPy array contains an instance of a particular row we can use the tolist(~) method.

To compute the Euclidean distance between two vectors in NumPy, use the np.linalg.norm(~) method.

To concatenate horizontally, we can use either use the np.concatenate(~) or np.hstack(~) method. To concatenate 1D Numpy arrays vertically, we can use the np.vstack(~) method.

To convert a NumPy array of strings to lowercase, use the np.char.lower(~) method.

To convert the type of a NumPy array to string, use astype(str).

To create a copy of a Numpy array, use its copy() method.

The major benefit of Numpy arrays over Python lists is when you need to perform calculations over all the items contained within the array / list.

The key difference is that array_equal returns True when the shape of the arrays exactly match, whereas array_equiv will also return True if the one array can be broadcasted to take on the same shape.

What differentiates mod(~) from Numpy's fmod(~) is confusingly not whether or not one is for floating numbers; they are both capable of parsing floating numbers. The defining difference is how they handle negative numbers.

The key difference is that Numpy's power(~) method uses the same data-type as the input array to perform the calculation; if your input array only contains integers, then the returned result will also be of type int. On the other hand, float_power(~) always uses float64 for maximum precision.

We can find the index of the largest value in a NumPy Array using the argmax(~) function of the Numpy array or using the argmax(~) static function of the Numpy module.

It is possible to find the index of smallest value in a Numpy array using the argmin(~) function.

To find the closest value in the Numpy array to a certain value we can leverage the following custom function.

To find the most frequent value in a NumPy array we can use the bincount(~) method together with argmax(~).

We can reduce a n-dimensional Numpy array to 1D using either the flatten() or ravel() method available to all Numpy arrays. We first demonstrate their usages, and subsequently their differences.

To get pi as a constant in NumPy use np.pi.

To extract elements from a two dimensional array using two dimensional array of indices in NumPy, use x[np.arange(x.shape[0])[:, None], idx].

To get the indices of N maximum values in NumPy in an optimal way, use the argpartition(~) method. The argpartition(~) method allows you to select an index by which to partition, and ensures that all values at indices smaller than the value at this index appear before it, and all values at indices larger than this index value appear after it.

To get the indices of N miniumum values in NumPy in an optimal way, use the argpartition(~) method. The argpartition(~) method allows you to select an index by which to partition, and ensures that all values at indices smaller than the value at this index appear before it, and all values at indices larger than this index value appear after it.

To get the number of columns of a 2D Numpy array, use its shape property.

To get the number of non-zero elements in a NumPy array use the count_nonzero(~) method.

To get the number of rows of a 2D Numpy array, use its shape property.

To initialize a Numpy array of zeros, use Numpy's ones(~) method.

To initialise a Numpy array of zeros, use Numpy's zeros(~) method.

To initialize an identity matrix in Numpy, use the np.eye(~) method.

To limit array values to a certain range in NumPy use the clip(~) method. Using this method, values within the array that fall outside the specified range will be replaced with the provided minimum and maximum values.

Linear regression, in essence, is about computing the line of best fit given some data points. We can use NumPy's polyfit(~) method to find this line of best fit easily.

Whether the full or truncated NumPy array is printed can be controlled using the threshold parameter of set_printoptions(). By default, threshold=1000.

We can use Numpy's set_printoptions(~) method to print large numpy arrays without truncation.

To remove rows containing NaN in a NumPy array, we can use a combination of the isnan(~) and any(~) methods.

To reverse a NumPy array use the flipud(~) method for one-dimensional NumPy arrays and fliplr(~) for multi-dimensional NumPy arrays.

To save a NumPy array to a file, we can use the savetext(~) method.

The shape property of a NumPy array is a tuple that tells us the size of the each nested array. This property is most commonly used to access the number of rows and columns of a 2D NumPy array.

To sort the value of one array according to another in Python use the np.argsort(~) method.

To suppress scientific notation when printing NumPy arrays, use np.set_printoptions(suppress=True).