Programming with PsychoPy, 7 Data Import and Export

So far, we have focused on stimuli presentation and responding to these stimuli. We let PsychoPy create the stimuli and put the results on the screen. But we need to record the data in external data files and sometimes we have already existing stimuli lists, we want to use for our experiment. For these functions we need data import and export.

Export

For exporting data we can use the PsychoPy experiment handler or the underlying Python file handling system. We choose the second since it is simpler to use and offers more freedom.

Example: saving your data

#!/usr/bin/env python
import csv

## Experiment Section, collect data here
data = [1, 0, 1, 1, 1, 0, 0, 1, 1]

## Closing Section
datafile = open("data.csv", "wb")
writer = csv.writer(datafile, delimiter=";")
for i, row in enumerate(data):
	writer.writerow([i, row])
datafile.close()

Saving data in Python works in two steps. What you have to do first is to define an output file where Python can direct output to. You do this by using the open() function. Second, you have to make a csv.writer that send your list of data to the output file every time you want to write to it. The term csv stand for comma separated value. It can really do more that just that, but we will see about that in the exercise.

For open(), you need the file name (i.e. how to call the file on the hard disk) as argument and a second argument that describes how to open the file. We use "wb" for writing en "rb" for reading. There are two more functions needed, writerow() and close(). With writerow() from csv.writer object you add a line of comma separated values to your output file on the hard disk, with close() from the file object, you close your file.

Importing

Importing or reading in data is almost as easy as exporting and works in a comparable manner. Imagine you want to present a square of a certain size and color on the screen. The experiment consists of four trials, but this could be a different set of trials for different groups of testees. Let's says that you have a file for each experiment and these files have one line for each trial. The value in the first column is the size of the square and the value in the second column is the time it should be on the screen. The columns are separated by a semicolon:

3.3;"red"
1.1;"green"
4.4;"red"
2.2;"green"

The experiment code code look like his:

#!/usr/bin/env python
import csv

## Setup section, read experiment variables size and color from file
size = []
color = []
datafile = open("stimuli.csv", "rb")
reader = csv.reader(datafile, delimiter=";")
for row in reader:
	size.append(float(row[0]))
	color.append(row[1])
datafile.close()

## Experiment Section, use experiment variables here
for trial in range(len(size)):
	print("size = {}, color = {}".format(size[trial], color[trial]))

First we make the empty lists to store size and color of the stimulus. Then we make a file object and connect it to a csvreader. The reader wants to know the precise format of the file, such as the delimiter used. Then we read every line of the file. The first field is appended to the list of sizes. The second row to the list of colors. This example does not show a real experiment section. It just prints the values to the PsychoPy output in the bottom of your window.

You do not need the csv.reader directly in your experiment. my.py as given in the previous lesson contains functions that can do this for you.

Assignment 7: Red-Green

Type the following experiment:

#!/usr/bin/env python
# -*- coding: utf-8 -*-

from psychopy import core, visual, event
import csv
 
## Setup section, read experiment variables from file
win = visual.Window([400,300], monitor="testMonitor", units="cm", fullscr=False)
stimuli = []
datafile = open("redgreen_stimuli.csv", "rb")
reader = csv.reader(datafile, delimiter=";")
for row in reader:
	if len(row)==2:         # ignore empty and incomplete lines
		size = float(row[0])  # the first element in the row converted to a floating point number
		color = row[1]        # the second element in the row
		stimulus = visual.Rect(win, width=size, height=size)
		stimulus.fillColor = color
		stimuli.append(stimulus)
datafile.close()
 
## Experiment Section, use experiment variables here
for stimulus in stimuli:
	stimulus.draw()
	win.flip()
	core.wait(1.000)

## Closing Section
win.close()
core.quit()

Run this experiment. You can make the stimuli file with the Psychopy editor or with Excel/ Libreoffice Calc/ ... . Choose Save as CSV (Comma delimited) if you use the latter.
Change the order of the presented stimuli in a random way. Use the shuffle function from the Python random module for this.
Change the experiment so that the block is shown until a key is pressed. Store the reaction time and write it to the screen.
Make sure that the participant has to respond within 1 second. If there is no response within one second write to the screen ‘Please respond faster’.
Extend your program that it writes the results to a file that contains original trial number, box size, box color and reaction time for each trial.

Warning for use with Microsoft Excel

Microsoft Excel uses the List separator from the Region and Language settings as delimiter when saving a csv-file, even though in the Save File dialog it says CSV (Comma delimited) (*.csv). This is a long standing problem in Microsoft Excel that especially affects European users, since in European version of Microsoft Windows the standard list separator is set to semi-colon (;). Since the interpretation of the comma (and the period) as decimal marker (and therefore not as list separator) is considered standard in the ISO 80 000 International System (pdf 5.3.4) we assume in this tutorial that the list separator is set to semi-colon.

Unlike most other programs, Microsoft Excel does not support the UTF-8 character set in csv-files. Keep in mind that you cannot use non-ascii characters (χαρακτήρες) if you make your experiment file in Excel.

Continue with the next lesson