Python Interpreters and Interactive Sessions

When we execute (i.e., run) a Python program, an interpreter translates the program code into an efficient intermediate representation and carries out the corresponding instructions, which are ultimately represented in the lowest-level computer language called machine code. Instructions in machine code can be given directly to the processor and thereby executed.

An interpreter is a program that translates another program line-by-line. There is another way of translating a program (written in a programming language) to machine code—compiling. A compiler takes the entire program at once and translates it into a highly optimized machine code program, ready for execution. An interpreter cannot optimize a program as much as can a compiler, but there are advantages to using an interpreter, including that programs can be run interactively.

The official Python interpreter CPython was installed to our development environment in section 1.4. The basic way to run a Python program hello.py is in a terminal window with the command

python hello.py

Here the interpreter program python is called to interpret hello.py and the results are printed to the terminal. Programs written in a file like hello.py are called scripts.

Another way to run a Python script is within an interactive session (i.e., interactive shell, read-evaluate-print loop (REPL), or kernel) that runs lines of code as they are entered by the programmer. There are multiple programs that provide interactive Python sessions, including the standard one provided by the CPython distribution and invoked in a termal window with the command python (without a script given). This command causes the terminal to start an interactive session with a prompt that appears similar to the following:

$ python
Python 3.<specific version number> | <additional information>
Type "help", "copyright", "credits" or "license" for more information.
>>> _

Python statements can be entered here and executed by pressing the key keyboard_return. For instance, a list can be created and sorted:

>>> l = [1.3, 0.8, 6.1, 3.9]
>>> l.sort()
>>> l
[0.8, 1.3, 3.9, 6.1]

Note that no explicit print() function call is necessary for objects to be printed in the interactive session. When a statement returns a value of probable interest, the session automatically prints it.

A more richly featured program for interactive Python sessions is IPython. The Spyder IDE in our development environment provides an IPython console (like a terminal) in the lower-right corner in the default layout.

In[1]: x = 4.29
In[2]: y = -38.1
In[3]: x**2 + x*y + y**2
Out[3]: 1306.5651

Within an IPython interactive session, a script can be run with the runfile() function. This is precisely what Spyder does when a script in the editor is run via the RunRun menu item, the button, or the F5 key. An advantage to this technique is that the variables defined in the script now enter the existing IPython interactive session, which allows us to play with them for further analytic exploration or for debugging purposes. Note however that the reverse is not the case: variables from the interactive session are not available to the script run within that session. This allows us to have confidence that a script developed with the use of runfile() in an interactive session can be run outside that session. Therefore, in this book, we need not be concerned about which method—direct calling of the interpreter via python or the use of runfile() within an interactive session—was used to run a Python script.