-
Notifications
You must be signed in to change notification settings - Fork 1
/
tuts0104-implementing-an-analysis.Rmd
161 lines (116 loc) · 6.49 KB
/
tuts0104-implementing-an-analysis.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
---
title: "Implementing an Analysis"
---
In this section we will add the implementation, that is the R code, to perform our t-test analysis.
In jamovi analyses, the implementation lives in the `.b.R` file, so if we look in our `ttest.b.R` file we see:
```{r eval=FALSE }
# This file is a generated template, your changes will not be overwritten
#' @rdname jamovi
#' @export
ttestClass <- R6::R6Class(
"ttestClass",
inherit = ttestBase,
private = list(
.run = function() {
# `self$data` contains the data
# `self$options` contains the options
# `self$results` contains the results object (to populate)
})
)
```
This is another file that our call to `jmvtools::create()` created. Now this may appear unfamiliar, and might not look like most of the R code you've written, but that's OK, you don't really need to know what's going on here. What *is* going on here is that the analysis is represented by an R6 class. For the curious, you can read more about R6 classes [here](https://cran.r-project.org/web/packages/R6/vignettes/Introduction.html), but all you really need to know is that you write your analysis in the `.run` function, and you can safely ignore the rest.
You'll also notice that the `.run()` function receives no arguments. We access the values that the user specified (either in the jamovi ui, or as arguments to the generated `ttestIS()` function) through `self`. Again, this may seem a little unfamiliar, but it is very straight forward.
As covered in the [previous section](tuts0103-creating-an-analysis.html), our t-test has four options (as defined in `ttest.a.yaml`), `dep`, `group`, `alt` and `varEq`, we can access the values for each of these in our analysis with:
- `self$options$dep`
- `self$options$group`
- `self$options$alt`
- `self$options$varEq`
Additionally, `ttest.a.yaml` defined the special `data` option, which means we can access the data provided by the user as a data frame (either the data loaded in jamovi, or the data passed as an argument to `ttestIS()` function in R), with:
```{r eval=FALSE }
self$data
```
Now we have access to the options, and access to the data, we can begin writing our analysis as follows:
```{r eval=FALSE }
ttestClass <- R6::R6Class("ttestClass",
inherit=ttestBase,
private=list(
.run=function() {
formula <- paste(self$options$dep, '~', self$options$group)
formula <- as.formula(formula)
results <- t.test(formula, self$data)
print(results)
})
)
```
First, we take the values of `self$options$dep` and `self$options$group`, which are both strings and assemble them into a formula. Then we can call the `t.test()` function passing in this formula, and the `self$data` data frame. Finally, we print the result.
Now this analysis will and does work; however when running in jamovi, the result of the print statement will appear at the terminal, rather than in the application's results area (where the user would like it). To remedy this, rather than simply printing the results, we assign the results to the analysis' results object. When run in an R session, the results will still be printed, but when run in jamovi, the results will appear in the results panel. We assign to the analysis' result object using (you guessed it), `self$results`. Our new function will now read:
```{r eval=FALSE }
ttestClass <- R6::R6Class("ttestClass",
inherit=ttestBase,
private=list(
.run=function() {
formula <- paste(self$options$dep, '~', self$options$group)
formula <- as.formula(formula)
results <- t.test(formula, self$data, var.equal=self$options$varEq)
self$results$text$setContent(results)
})
)
```
In this new function, we get the results element called `text` from `self$results`, and call `setContent()` with the results from the t-test. We'll cover results elements in greater depth in the next section, but for now this is all you need to know.
So now our analysis is implemented, it's time to install it and try it out. Install the module with the usual:
```{r eval=FALSE }
jmvtools::install()
```
Now open the `Tooth Growth` data set from the jamovi examples (File -> Examples -> Tooth Growth). Assign the `len` column to the `Dependent Variable`, and the `supp` column to the `Grouping Variable`. You should have something like the following:
{ width=1260px }
Similarly, we can install this module as an R package using the `devtools` package (not to be confused with `jmvtools`), and run the same analysis in an interactive R session:
```{r, eval=FALSE}
devtools::install()
library(SuperAwesome)
data(ToothGrowth)
ttest(data=ToothGrowth, dep='len', group='supp')
```
```
Independent Samples T-Test
Two Sample t-test
data: len by supp
t = 1.9153, df = 58, p-value = 0.06039
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
-0.1670064 7.5670064
sample estimates:
mean in group OJ mean in group VC
20.66333 16.96333
```
Before we continue, astute readers will have realised that assembling our formula with `paste` is problematic. If either column name has spaces or special characters, paste will produce a bad formula. For example, if the user specified a dependent variable called `the fish` --- the resultant formula would be `the fish~group`, and the call to `as.formula()` would fail:
```{r, eval=FALSE}
as.formula('the fish~group')
```
```
## Error in parse(text = x, keep.source = FALSE) :
## <text>:1:5: unexpected symbol
## 1: the fish
## ^
```
The names of the columns making up the formula need to be escaped, or quoted. Fortunately, `jmvcore` provides the function `constructFormula()`, which assembles simple formulas appropriately escaping column names:
```{r, eval=FALSE}
constructFormula('the fish', c('group'))
```
```
## [1] "`the fish`~group"
```
We can modify our analysis to use this instead:
```{r, eval=FALSE}
ttestISClass <- R6Class("ttestISClass",
inherit=ttestISBase,
private=list(
.run=function() {
formula <- constructFormula(self$options$dep, self$options$group)
formula <- as.formula(formula)
results <- t.test(formula, self$data)
self$results$text$setContent(results)
})
)
```
The `jmvcore` package contains many such useful functions. It would be worth checking them out.
Next: [Debugging an Analysis](tuts0105-debugging-an-analysis.html)