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()
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
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
Note
An individual object is an instance of a class. That is, a class is a set of code and specifications which defines what an object created from it can do. We create a specific object by instantiating it using the class as a recipe. This terminology perhaps won’t make sense until you see the more sophisticated PsychoPy classes being used.
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.
Integers and floats (floating point numbers) are just 2 different types of number. Integers don’t store anything after the decimal place.
If you’re using Python2 beware that dividing a pair of integers gives you back an integer:
print(1/3) # surprise!
print(1.0/3)
print(1/3.0)
In Python3 dividing integers returns a float if needed, and you can get the same in Python2 by adding this to your script:
from __future__ import division
Python has fantastic string handling options. Try these methods that are attached to strings:
a = 'hello world'
a.title()
a.split()
a.endswith('world')
len(a) # you'll probably use this function a lot
You can also combine strings in nice, simple ways:
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()
Often you need to fetch a subset of an object, like a string or a list.
Warning
If you’re used to Matlab or R then be warned: in Python the first element of an array or a list is zero, not one. This will catch you out sooner or later!!
>>> a = 'Nottingham'
>>> a[0]
>>> a[2:4]
>>> a[2:]
>>> a[:]
>>> a[-1]
>>> a[2:-2].upper()
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])
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/
Now you know about simple variables let’s store them in 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.
For storing things that have a defined order:
a = [30, 20, 10]
b = ['a', 1, 1.0]
b.append('blah')
a.append(3.0)
Slicing works just the same as with strings:
a[0] # remember, Python starts at zero
a[4] # so this won't work
a[-1] # this will
b[-1]
b[-1][-1]
Mathematical operators:
a + a # this might be a surprise (unlike in Matlab/R)
a + b
b * 3
For those who have come from Matlab backgrounds, these lists might look like Matlab matrices, but they aren’t. These aren’t designed for mathematical operations. There is a similar object which is very much like Matlab matrices, which we’ll explore when we look at Data analysis
Other methods:
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?
At times you want to keep things with something that identifies what each element is. That’s where you’ll use a dict. These can be created in various ways:
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'])
Then you can access the contents in a similar way:
print(stim1['ori'])
print(stim2['fail']) # error?
Explore what some of the different dict methods do:
dir(stim1)
stim1.keys()
stim1.values()
To test if a dictionary has a particular key:
'word' in stim1 # True
'blue' in stim1 # False
Dictionaries are important in PsychoPy. They are often used to hold the information specifying parameters of a trial:
trial1 = {'distractor':True,
'image':'cat.jpg',
'duration':0.5}
NB. if a line ends in a comma you can break the line without breaking the code
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!
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')
Note
upper() is a method that all strings have. Let’s find out what else they have by using the dir() function. Add print(dir(name)) to the last line.
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')
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)
Note
Note that in the shell if the command returns a value and you didn’t provide anything to receive/store that value then it gets printed to the screen instead (this is not the case for scripts run from the editor).
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)
>>> from psychopy import visual, core, data, event, sound, gui
>>> from psychopy.constants import * # like STARTED, FINISHED
A general strategy for approaching errors is:
Look up the line that caused the error!
- Did you type everything correctly?
- Do all your parenthesis, brackets, and quotes match?
- Did you do something you shouldn’t have?
- Did you call for a class that doesn’t exist, wasn’t imported, or hasn’t been set yet?
- Is something not what you thought it was?
- Was a value possibly redefined when you “weren’t looking” or did a variable have a misleading name, like a number called “subject_name”?
The variable is not defined (yet):
>>> myVariable = 2
>>> print(myvariable)
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
NameError: name 'myvariable' is not defined
An error as output from PsychoPy:
## 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
The variables are of the wrong type for what you tried to do to them:
>>> subject = 'Emma'
>>> trialNum = 2
>>> print(subject + trialNum)
TypeError: cannot concatenate 'str' and 'int' objects
Revision: How do you concatenate a string and an integer?
Your statements don’t follow the Python syntax. Because of this, the error message can’t give helpful details except where the error occurred (via the line number and caret ^ symbol.
>>> my Variable = 2
File "<stdin>", line 1
my Variable = 2
^
SyntaxError: invalid syntax
Another syntax error example:
>>> for i in range(10)
>>> print(i ** 2)
for i in range(10)
^
SyntaxError: invalid syntax
You tried to access an element of a list using an index which is out of bounds.
>>> a = [10, 20, 30]
>>> a[3]
IndexError: list index out of range
You tried to access an entry in a dictionary using a key that doesn’t exist.
>>> details = {'name': 'jonas'}
>>> print(details['age'])
KeyError: 'age'
- Try this flow chart: http://i.imgur.com/WRuJV6r.png
- For the pros, see http://stackoverflow.com/questions/1011431/common-pitfalls-in-python
