IO with NumPy - use Technical Quotient

1. Remember

What is the default dtype when using np.genfromtxt() without specifying any dtype argument?

str
None

Incorrect; None leads to automatic data type determination, not the default behavior.

int
float

Correct; the default dtype for np.genfromtxt() is float unless otherwise specified.

2. Understand

How does specifying dtype=None in np.genfromtxt() affect the reading of mixed data types from a file?

The output is a 2D array with homogeneous data type.
Data types must be manually specified after reading.
Function iteratively infers column data types automatically.

Correct; with dtype=None, np.genfromtxt() automatically determines the data type for each column based on the data.

All data is converted to strings.

3. Understand

What is the result of using a sequence of types, such as dtype=(int, float, float), when reading data with np.genfromtxt()?

The output will be a 1D array with a structured dtype.
Each column in the output array will have the same data type.
The output will be a 2D array with mixed data types.
A dictionary must be provided to specify data types for each column.

4. Apply

You have a CSV file data.csv with three columns: an integer, a floating-point number, and a string. You want to read this file into a NumPy array, ensuring each column has the correct data type. Which dtype specification should you use with np.genfromtxt()?

np.genfromtxt('data.csv', delimiter=',', dtype='i4,f8,|U3')

Incorrect; the 'U3' may not be sufficient for all strings in the third column, so the first option is more reliably correct for general cases.

np.genfromtxt('data.csv', delimiter=',', dtype=None)

Incorrect; although dtype=None would automatically determine the data type for each column, it is slower than specifying explicitly.

np.genfromtxt('data.csv', delimiter=',', dtype=(int, float, 'U10'))

Correct; dtype=(int, float, 'U10') explicitly specifies the data type for each column, matching the requirements.

np.genfromtxt('data.csv', delimiter=',', dtype='float')

Incorrect; dtype='float' would attempt to convert all data to floating-point numbers, which is not suitable for the string column.

1. Remember

2. Understand

3. Understand

4. Apply

Leave a ReplyCancel Reply