Tutorial 3 - Visualise data with
ggplot2
Author: Matthew J. Crossley
Last update: 02 April, 2024
Learning objectives
Learn how to wrange real data into a
data.tableand plot it usingggplot.Get lots of hands-on practice with
Randdata.tableandggplot.In particular, practice using the following functions:
ggplot()geom_point()geom_line()geom_bar()geom_hist()geom_violin()geom_boxplot()
Get some experiment data
Complete the experiment located at the following address:
https://run.pavlovia.org/demos/simplertt/Be sure to enter your name or some other ID that you will remember and can be easily searched for.
Download your data in a
.csvfile by taking the following steps:Go here:
https://gitlab.pavlovia.org/demos/simplerttRead about the experiment you just participated in by scrolling to the bottom and reading the
README.mdfile.Click the
datafolder to land at the following address:https://gitlab.pavlovia.org/demos/simplertt/tree/master/dataNear the top right of the page, click the
Find filebutton, and search for a file containing the unique ID that you entered at the beginning of the experiment.Click on the
.csvfile that pops up.Finally, click the
downloadbutton to download your.csvfile to your local machine.
Analyse your data using data.table
- Load the
data.tablelibrary andrmto be sure your are starting with a clean work space.
library(data.table)
library(ggplot2)
rm(list = ls())- Load the data into a
data.tableusing thefreadfunction from thedata.tablelibrary.
# You need to replace the path I use here with a path that
# points to wherever you have your data stored.
d <- fread('https://crossley.github.io/book_stats/data/real_data/simpleRTT/data_mjc_simpleRTT_2022-02-26_05h13.22.csv')We can see that
dhas quite a few columns by printing it to the console (which I won’t do here because the output is very ugly and I want to keep things clean). You, however, should do it, so you can see what I mean.When
data.tableobjects have lots of columns,strcan be a good summary function to use for basic inspection.
str(d)
## Classes 'data.table' and 'data.frame': 28 obs. of 26 variables:
## $ response.keys : chr "space" "space" "space" "space" ...
## $ response.corr : int 1 1 1 1 1 1 1 1 1 1 ...
## $ response.rt : num 0.306 0.273 0.283 0.317 0.3 0.286 0.272 0.3 0.308 0.284 ...
## $ mouseResp.x : num 0.228 0.228 0.228 0.228 0.228 ...
## $ mouseResp.y : num -0.334 -0.334 -0.334 -0.334 -0.334 ...
## $ mouseResp.leftButton : int 0 0 0 0 0 0 0 0 0 0 ...
## $ mouseResp.midButton : int 0 0 0 0 0 0 0 0 0 0 ...
## $ mouseResp.rightButton : int 0 0 0 0 0 0 0 0 0 0 ...
## $ practiceTrials.thisRepN : int 0 0 0 0 0 0 0 0 NA NA ...
## $ practiceTrials.thisTrialN: int 0 1 2 3 4 5 6 7 NA NA ...
## $ practiceTrials.thisN : int 0 1 2 3 4 5 6 7 NA NA ...
## $ practiceTrials.thisIndex : int 1 7 3 2 5 6 4 0 NA NA ...
## $ practiceTrials.ran : int 1 1 1 1 1 1 1 1 NA NA ...
## $ isi : num 1.22 2.56 1.67 1.44 2.11 ...
## $ participant : chr "mjc" "mjc" "mjc" "mjc" ...
## $ session : int 1 1 1 1 1 1 1 1 1 1 ...
## $ date : chr "2022-02-26_05h13.22" "2022-02-26_05h13.22" "2022-02-26_05h13.22" "2022-02-26_05h13.22" ...
## $ expName : chr "simpleRTT" "simpleRTT" "simpleRTT" "simpleRTT" ...
## $ psychopyVersion : chr "2021.2.3" "2021.2.3" "2021.2.3" "2021.2.3" ...
## $ OS : chr "MacIntel" "MacIntel" "MacIntel" "MacIntel" ...
## $ frameRate : num 58.8 58.8 58.8 58.8 58.8 ...
## $ mainTrials.thisRepN : int NA NA NA NA NA NA NA NA 0 0 ...
## $ mainTrials.thisTrialN : int NA NA NA NA NA NA NA NA 0 1 ...
## $ mainTrials.thisN : int NA NA NA NA NA NA NA NA 0 1 ...
## $ mainTrials.thisIndex : int NA NA NA NA NA NA NA NA 5 8 ...
## $ mainTrials.ran : int NA NA NA NA NA NA NA NA 1 1 ...
## - attr(*, ".internal.selfref")=<externalptr>It’s certainly difficult to know what all of these columns encode. This is something you will get used to as you build and run your own experiments (e.g., as you will in later COGS units). For now, I’ll just tell you. The data contains a row for every trial completed, including practice trials.
You can tell which rows correspond to practice trials and which correspond to experiment trials by examining the columns named
practiceTrials.thisNandmainTrials.thisN.
d[, .(practiceTrials.thisN, mainTrials.thisN)]
## practiceTrials.thisN mainTrials.thisN
## <int> <int>
## 1: 0 NA
## 2: 1 NA
## 3: 2 NA
## 4: 3 NA
## 5: 4 NA
## 6: 5 NA
## 7: 6 NA
## 8: 7 NA
## 9: NA 0
## 10: NA 1
## 11: NA 2
## 12: NA 3
## 13: NA 4
## 14: NA 5
## 15: NA 6
## 16: NA 7
## 17: NA 8
## 18: NA 9
## 19: NA 10
## 20: NA 11
## 21: NA 12
## 22: NA 13
## 23: NA 14
## 24: NA 15
## 25: NA 16
## 26: NA 17
## 27: NA 18
## 28: NA 19
## practiceTrials.thisN mainTrials.thisNWe can see that
mainTrials.thisNisNAduring practice andpracticeTrials.thisNisNAduring the main experiment.We can also see that our reaction time per trial is stored in a column named
response.rtand the inter-stimulus interval (isi) is stored in a column namedisi.response.rtis our main dependent variable of interest, andisiis an independent variable.A dependent variable is a variable or outcome not controlled by the experimenter but rather observed in the experiment.
An independent variable is a variable or experiment factor controlled by the experimenter that may influence observations from a dependent variable.
You can also see that both
response.RTandisiare continuous (as opposed to discrete, or categorical etc.) observations numbers bounded between zero and positive infinity.We will talk about these key terms in more depth coming up in lecture.
Okay, we are now equipped to pull out just the rows and columns that we need for a simple exploration of our performance.
# We begin by just looking and the main trials
d_main <- d[!is.na(mainTrials.thisN), .(response.rt, isi)]
# We don't have enough data points for a histogram to look
# very good, but it is generally a good place to start so
# you get a feel for the shape of your data.
ggplot(d_main, aes(x=response.rt)) +
geom_histogram(bins=30)
# The histogram shows that between 0.2 and 0.3 ms our
# observations seem a bit bell-shaped. However, we can
# clearly see that a response time is never negative and is
# sometimes quite far away from the peak of the bell shape
# (where most of the observations are located).
# From the histogram, you can eye-ball a reasonable guess
# about the central tendency and spread of this sample, but
# it would also be very reasonable to dig a bit deeper with
# some additional plots.
ggplot(d_main, aes(x=0, y=response.rt)) +
geom_boxplot() +
geom_point() +
xlim(-1, 1)
# Finally, we report basic descriptive statistics as actual
# numbers for the main experiment
d_main[, .(rt_mean=mean(response.rt),
rt_median=median(response.rt),
rt_sd=sd(response.rt))]
## rt_mean rt_median rt_sd
## <num> <num> <num>
## 1: 0.31745 0.269 0.1530361
# Report basic descriptive statistics for the main
# experiment using `chaining` to avoid creating a new
# data.table
d[!is.na(mainTrials.thisN), .(response.rt, isi)][, .(rt_mean=mean(response.rt),
rt_median=median(response.rt),
rt_sd=sd(response.rt))]
## rt_mean rt_median rt_sd
## <num> <num> <num>
## 1: 0.31745 0.269 0.1530361
# You should check these numbers against the plots you've
# made to ensure they are all consistent with each other.- Report basic descriptive statistics for practice and for the main
experiment at the same time. To do this, it will be very helpful to have
a new single column that indicates whether a particular row id
dcorresponds to a practice trial or to a main trial.
d[, phase := "Unknown"]
d[!is.na(practiceTrials.thisN), phase := "practice"]
d[!is.na(mainTrials.thisN), phase := "main"]
d[, .(phase, response.rt)]
## phase response.rt
## <char> <num>
## 1: practice 0.306
## 2: practice 0.273
## 3: practice 0.283
## 4: practice 0.317
## 5: practice 0.300
## 6: practice 0.286
## 7: practice 0.272
## 8: practice 0.300
## 9: main 0.308
## 10: main 0.284
## 11: main 0.322
## 12: main 0.267
## 13: main 0.284
## 14: main 0.266
## 15: main 0.271
## 16: main 0.251
## 17: main 0.241
## 18: main 0.267
## 19: main 0.302
## 20: main 0.260
## 21: main 0.250
## 22: main 0.300
## 23: main 0.334
## 24: main 0.266
## 25: main 0.236
## 26: main 0.440
## 27: main 0.939
## 28: main 0.261
## phase response.rt- The reason having this new column helps so much is because we can
now use it to
groupdescriptive statistics as follows:
d[, .(rt_mean=mean(response.rt),
rt_median=median(response.rt),
rt_sd=sd(response.rt)),
.(phase)]
## phase rt_mean rt_median rt_sd
## <char> <num> <num> <num>
## 1: practice 0.292125 0.293 0.01615494
## 2: main 0.317450 0.269 0.15303611- It also facilitates plotting comparisons as follows:
ggplot(d, aes(x=phase, y=response.rt)) +
geom_boxplot()
- From here, we can ask if there is any relationship between our dependent and our independent variables. When looking for relationships between two variables, one of the first things we usually do is plot one of them on the x-axis and the other of them on the y-axis.
ggplot(d, aes(x=isi, y=response.rt)) +
geom_point()
- No obvious relationship jumps out to my eye. You?
Work through these practice exercises
It’s a good idea to work through these on your own, but if you get very stuck, solutions can be found here
These practice problems use the
diamondsdata set that is loaded automatically when you loadggplot2. By default it exists as adata.frame. You can convert it to adata.tablewith the following line:For each practice problem, replicate each
ggplotfigure anddata.tableoutput shown below exactly.
1.
## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
## cut mean_carat range_carat
## <ord> <num> <num>
## 1: Ideal 0.702837 3.32.
## cut clarity N
## <ord> <ord> <int>
## 1: Ideal SI2 2598
## 2: Premium SI1 3575
## 3: Good VS1 648
## 4: Premium VS2 3357
## 5: Good SI2 1081
## 6: Very Good VVS2 1235
## 7: Very Good VVS1 789
## 8: Very Good SI1 3240
## 9: Fair VS2 261
## 10: Very Good VS1 1775
## 11: Good SI1 1560
## 12: Ideal VS1 3589
## 13: Premium SI2 2949
## 14: Premium I1 205
## 15: Very Good VS2 2591
## 16: Premium VS1 1989
## 17: Ideal SI1 4282
## 18: Good VS2 978
## 19: Very Good SI2 2100
## 20: Ideal VVS2 2606
## 21: Ideal VVS1 2047
## 22: Premium VVS1 616
## 23: Good VVS1 186
## 24: Premium VVS2 870
## 25: Fair SI2 466
## 26: Ideal VS2 5071
## 27: Good VVS2 286
## 28: Very Good I1 84
## 29: Fair SI1 408
## 30: Ideal IF 1212
## 31: Fair I1 210
## 32: Premium IF 230
## 33: Fair VVS1 17
## 34: Very Good IF 268
## 35: Fair VS1 170
## 36: Ideal I1 146
## 37: Good I1 96
## 38: Fair IF 9
## 39: Fair VVS2 69
## 40: Good IF 71
## cut clarity N3.
## cut clarity median_cty range_cty
## <ord> <ord> <num> <int>
## 1: Ideal SI2 599.0 670
## 2: Premium SI1 709.0 672
## 3: Good VS1 616.0 662
## 4: Premium VS2 734.0 665
## 5: Good SI2 593.0 662
## 6: Very Good VVS2 605.0 652
## 7: Very Good VVS1 642.0 659
## 8: Very Good SI1 660.5 656
## 9: Fair VS2 813.0 659
## 10: Very Good VS1 640.0 660
## 11: Good SI1 606.0 655
## 12: Ideal VS1 716.0 659
## 13: Premium SI2 631.0 652
## 14: Premium I1 835.0 608
## 15: Very Good VS2 650.0 644
## 16: Premium VS1 734.0 644
## 17: Ideal SI1 660.0 642
## 18: Good VS2 643.0 630
## 19: Very Good SI2 585.0 614
## 20: Ideal VVS2 775.5 586
## 21: Ideal VVS1 815.0 583
## 22: Premium VVS1 803.0 583
## 23: Good VVS1 724.5 591
## 24: Premium VVS2 796.0 554
## 25: Ideal VS2 734.0 632
## 26: Very Good I1 720.0 427
## 27: Ideal IF 886.0 531
## 28: Very Good IF 865.0 630
## 29: Fair VS1 735.0 611
## 30: Good VVS2 645.0 621
## 31: Premium IF 895.0 455
## 32: Fair SI1 775.0 498
## 33: Fair I1 893.0 408
## 34: Fair VVS2 750.0 628
## 35: Ideal I1 530.0 558
## 36: Good IF 827.0 496
## 37: Good I1 497.0 570
## 38: Fair SI2 871.0 439
## 39: Fair VVS1 790.0 135
## cut clarity median_cty range_cty4.
