Variables and common types¶

Assigning variables¶

You’ve already seen variables being assigned using =. You can also assign multiple things at once. Type these in and then check to see what each variable looks like afterwards:

a = b = 2

a, b = 2, 3 # try what happens if the number of vals doesn't match

myString = 'Hello World'

firstWord, secondWord = myString.split() # also try splitting by 'o'

listOfTwoWords = myString.split()

Comparing variables¶

In some programming languages, the ‘=’ sign is used both to assign a value to a variable and to compare variables for equality.

In Python, we use ‘==’ to test if two things are equal:

>>> a = 2 # make a equal 2
>>> b = a # make b equal a
>>> b == a # compare b to a
True

Types of variable¶

In Python almost everything is an object. There are a number of built-in objects that are very widely used:

• integers and floats

• booleans

• strings (and unicodes)

• tuples and lists

• dicts

Objects can be much more complicated.

PsychoPy adds a bunch of other object types, such as:

Window # something to draw stimuli in

GratingStim # something psychophysicists love to draw in windows

DotStim # lots of random dots that move by themselves

Clock # control timing

ExperimentHandler # control trials & save data

ParallelPort # send signals to hardware like EEGs

You can find out what an object is using the type() function:

type('Hello')
type('Hellø')  # in Python3 a string is "unicode" format
type(5)
type(5.0)
type([1, 2, 3])
type(True)

Finding out the type can be essential if things look the same when printed but aren’t the same. Sprinkling print statements like this throughout your code is often a key part of debugging in Python/PsychoPy:

a = 5
b = '5'
print(a, b)
print(a == b)
print(type(a), type(b))

But there are a huge number of additional objects and you can make your own too.

print(1/3) # surprise!
print(1.0/3)
print(1/3.0)

from __future__ import division

a = 'hello world'
a.title()
a.split()
a.endswith('world')
len(a) # you'll probably use this function a lot

image = 'cat'
suffix = '.jpg'
fileName = image + suffix # you can add or 'concatenate' strings
text = image * 2 # yes, this is valid
text = image * 2.0 # but this isn't
image + image.upper()

Slicing¶

Often you need to fetch a subset of an object, like a string or a list.

>>> a = 'Nottingham'
>>> a[0]
>>> a[2:4]
>>> a[2:]
>>> a[:]
>>> a[-1]
>>> a[2:-2].upper()

Converting (aka coercing)¶

You can convert between these different types of objects where they make sense:

a = int(1.5)
b = int('1')
c = str(1.5)
d = float(1)

but not where they don’t:

float('f1')
int([1,2])

Formatted strings¶

Sometimes you need to combine numbers and strings. Imagine I wanted to make a filename to save my data. Maybe in my script I had a variable to store my subject name and another to store a stimulus attribute which was 10, 50, 100, 200 on different runs. I might try and save the data filename like this:

subj = 'jwp'
cond = 50.0
filename = subj + cond + ".txt"

You get an error because cond is a number and your trying to add it to a pari of strings (subj and “.txt”) and Python doesn’t know what way you want them combined.

You could convert cond into a string and have no error:

filename = subj + str(cond) + ".txt"

Instead, we can use a formatted string or “f-string” in Python3 (we’ll mention the Python2 options later):

filename = f"{subj}{cond}.txt"

You can also specify the format of numbers. Try some of these:

f"{subj}{cond}.txt"  # default format for that object type
f"{subj}{cond:.3f}.txt"
f"{subj}{cond:.0f}.txt"
f"{subj}{cond:6.0f}.txt"
f"{subj}{cond:06.0f}.txt"
f"{subj}{cond:+06.0f}.txt"

These would often be used in feedback messages too:

rt = 0.63445345
msg = f"Well done {subj}! Your reaction time was {rt:.3f}"

You can even run simple code snippets inside these formatted strings!:

f"{subj.upper()}.txt"

The above system of formatting is only for Python3.6+ but there are other systems too (use these if you need Python2 compatibility).

Like the new f-strings:

filename = "{}{:03}.txt".format(subj, cond)

Like older C-style formats:

filename = "%s%03i.txt" %(subj, cond)

For more see on those two see: https://pyformat.info/

Containers¶

Very often you need variables that store more than one value and keep them organised in some way. The two most common are lists and dictionaries.

a = [30, 20, 10]
b = ['a', 1, 1.0]
b.append('blah')
a.append(3.0)

a[0] # remember, Python starts at zero
a[4] # so this won't work
a[-1] # this will
b[-1]
b[-1][-1]

a + a # this might be a surprise (unlike in Matlab/R)
a + b
b * 3

print(dir([])) # go and explore some of the other methods of lists
a.pop() # pull off the last element of a
a.append(b) # add all of b as the last element of a
a.extend(b) # add all of the elements of b individually
a.index(30) # where is the value 30?

stim1 = {'word':'red','ori':90,'duration':0.5}

# or just create it and add the entries afterwards:
stim2 = {}
stim2['word'] = 'blue'
stim2['ori'] = 90
stim2['duration'] = 0.3
print(stim1['word'])

print(stim1['ori'])
print(stim2['fail']) # error?

dir(stim1)
stim1.keys()
stim1.values()

'word' in stim1  # True
'blue' in stim1  # False

trial1 = {'distractor':True,
          'image':'cat.jpg',
          'duration':0.5}

Nesting objects within each other¶

Often containers are nested within each other. You might well have a list of dicts, or a dict containing lists etc.:

#a list of dicts
stimuli = [stim1, stim2, stim3]
stimuli[0]['word'] # this is stim1 because we start at zero!!
thisStimulus = stimuli[2]
thisStimulus == stim3

or a list of lists:

coordinates=[[0,0], [2,3], [8,0]]
responses = [ [1, 1, 0, 0],
    [1,1,1,0],
    [0,1,1,1]]
print(responses)
print(responses[2][3]) #the 4th entry of 3rd list (STARTS AT ZERO)

#or we could have done this
responses = []
cond1 = [1, 1, 0, 0]
responses.append(cond1) #etc.

You can nest objects as deeply as you like. The limit is your own brain being able to keep track of what you’re doing!

Indentation¶

One of the unusual things in Python is that indentation (whitespace) is actually important. Try to use a genuine programmer’s text editor and set it to insert spaces in place of tabs (it’s hard to spot errors when you have a mixture of tabs and spaces). Many editors, will try to help you get indentation right:

if response == 'y':
   print('Yes')
else:
   print('No')
print('Whatever')

Type the following into the editor after your other text:

name = 'Jessica'
for thisLetter in name:
    print(thisLetter)
    print(thisLetter.upper())
    print('done')

Now, that probably didn’t do what you expected. In Python the code that is included as part of the for-loop is indicated by the level of indentation, so print('done') was repeated for each repeat.

Select the last few lines of code and press Ctrl-[ to get this:

for thisLetter in name:
    print(thisLetter)
print(thisLetter.upper())
print('done')
print(dir(name))

Now the code will print each of the letters in their lower case. Then the loop ends. thisLetter still exists but it isn’t changing any more. It gets printed just once in upper case, followed by the other commands.

print can print multiple objects at once (if you insert commas), and you can suppress the line endings. With old-style print:

for thisLetter in name:
    print thisLetter, thisLetter.upper(),
print 'end of the loop'
print 'done'

Or new-style (Python3 or using from __future__ import print):

for thisLetter in name:
    print(thisLetter, thisLetter.upper(), end="")
print('end of the loop')
print('done')

Importing modules:¶

Python functions beyond its basic set are organised into modules and packages (PsychoPy is a set of such modules). We need to explicitly import such packages to be able to access their functions:

>>> import os # handy system and path functions

You can find out what’s in a module using the function dir():

>>> dir(os)

Import statement variants¶

Having functions in modules allows us to avoid name space collisions, such as functions that have the same name. Different import styles allow different naming:

import numpy # access like:  numpy.sin(0)
import numpy as np # np.sin(0)
from numpy import sin, cos, tan # sin(0)
from numpy import * # often frowned upon

from numpy.random import random, randint, shuffle

# can now say shuffle(mylist) rather than
# numpy.random.shuffle(myList)

PsychoPy imports:¶

>>> from psychopy import visual, core, data, event, sound, gui

>>> from psychopy.constants import * # like STARTED, FINISHED

Common mistakes & error messages¶

To follow is a list of simple, common mistakes. Remember:

• every character counts. A single typo/omission causes the script to crash.

• the script runs from the top and down. Things not defined yet cannot be referenced (used).

Python is helpful in letting you debug errors.

• in every error, the top line tells you the line at which the error occurred. You probably mis-typed something.

• in every error, the bottom line tells you the type of error.

>>> myVariable = 2
>>> print(myvariable)

Traceback (most recent call last):
File "<stdin>", line 1, in <module>
NameError: name 'myvariable' is not defined

## Running: /Users/michael/Desktop/PsychoPy test/test_lastrun.py ###

Traceback (most recent call last):

File "/Users/michael/Desktop/PsychoPy test/test_lastrun.py",

line 96, in <module>

print(aVariableThatIHaventDefinedYet)

NameError: name 'aVariableThatIHaventDefinedYet' is not defined

>>> subject = 'Emma'
>>> trialNum = 2
>>> print(subject + trialNum)
TypeError: cannot concatenate 'str' and 'int' objects

>>> my Variable = 2
File "<stdin>", line 1
my Variable = 2
     ^
SyntaxError: invalid syntax

>>> for i in range(10)
>>>    print(i ** 2)

for i in range(10)
                 ^
SyntaxError: invalid syntax

>>> a = [10, 20, 30]
>>> a[3]
IndexError: list index out of range

>>> details = {'name': 'jonas'}
>>> print(details['age'])
KeyError: 'age'