-
-
Notifications
You must be signed in to change notification settings - Fork 522
/
15-parameterized.Rmd
190 lines (144 loc) · 7.7 KB
/
15-parameterized.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
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
# (PART) Other Topics {-}
# Parameterized reports
One of the many benefits of working with R Markdown is that you can reproduce analysis at the click of a button. This makes it very easy to update any work and alter any input parameters within the report. Parameterized reports\index{parameterized reports} extend this one step further, and allow users to specify one or more parameters to customize the analysis. This is useful if you want to create a report template that can be reused across multiple similar scenarios. Examples may include:
- Showing results for a specific geographic location.
- Running a report that covers a specific time period.
- Running a single analysis multiple times for different assumptions.
- Controlling the behavior of **knitr** (e.g., specify if you want the code to be displayed or not).
In this chapter, we discuss the use of parameterized reports, and explain how we can interactively define the parameters to compile the results.
## Declaring parameters {#params-declare}
Parameters are specified using the `params` field within the YAML section. We can specify one or more parameters with each item on a new line. As an example:
```yaml
---
title: My Document
output: html_document
params:
year: 2018
region: Europe
printcode: TRUE
data: file.csv
---
```
All standard R types that can be parsed by `yaml::yaml.load()` can be included as parameters, including `character`, `numeric`, `integer`, and `logical` types. We can also use R objects by including `!r` before R expressions. For example, we could include the current date with the following R code:
```yaml
---
title: My Document
output: html_document
params:
date: !r Sys.Date()
---
```
Any R expressions included within the parameters are executed before any code in the document, therefore any package dependencies must be explicitly stated using the `package::function` notation (e.g., `!r lubridate::today()`), even if the package is loaded later in the Rmd document.
## Using parameters {#params-use}
You can access the parameters within the knitting environment and the R console in RStudio.^[Parameters will not be available immediately after loading the file, but require any line of the report to be executed first.] The values are contained within a read-only list called `params`. In the previous example, the parameters can be accessed as follows:
```r
params$year
params$region
```
Parameters can also be used to control the behavior of **knitr**. For example, the **knitr** chunk option `echo` controls whether to display the program code, and we can set this option globally in a document via a parameter:
````markdown
---
params:
printcode: false # or set it to true
---
`r ''````{r, setup, include=FALSE}
# set this option in the first code chunk in the document
knitr::opts_chunk$set(echo = params$printcode)
```
````
## Knitting with parameters {#params-knit}
There are three ways in which a parameterized report can be knitted:
- Using the `Knit` button within RStudio.
- `rmarkdown::render()` with the `params` argument.
- Using an interactive user interface to input parameter values.
### The `Knit` button
By using the `Knit` button in RStudio or calling `rmarkdown::render()` function, the default values listed in the YAML metadata (if specified) will be used.
### Knit with custom parameters
Even if your document has the `params` field in the YAML metadata, you can actually override it by providing a custom list of parameter values to the function `rmarkdown::render()`. For example:
```{r, eval=FALSE, tidy=FALSE}
rmarkdown::render("MyDocument.Rmd", params = list(
year = 2017,
region = "Asia",
printcode = FALSE,
file = "file2.csv"
))
```
We do not have to explicitly state all parameters in the `params` argument. Any parameters not specified will default to the values specified in the YAML metadata. For example, this will only override the `region` parameter:
```{r, eval=FALSE, tidy=FALSE}
rmarkdown::render("MyDocument.Rmd", params = list(
region = "Asia"
))
```
You may want to integrate these changes into a function. Such a function could also be used to create an output file with a different filename for each of the different combination of parameters. In the following example, a new file `Report-region-year.pdf` is created for each set of parameters:
```{r, eval=FALSE, tidy=FALSE}
render_report = function(region, year) {
rmarkdown::render(
"MyDocument.Rmd", params = list(
region = region,
year = year
),
output_file = paste0("Report-", region, "-", year, ".pdf")
)
}
```
### The interactive user interface
We can use a graphical user interface (GUI) based on Shiny to interactively input the parameters of a report. The user interface can be called by either `rmarkdown::render("MyDocument.Rmd", params = "ask")` or clicking the drop-down menu behind the `Knit` button and choosing `Knit with Parameters` in RStudio. Figure \@ref(fig:params-input) shows the GUI of **rmarkdown** asking for inputting parameters.
```{r params-input, echo=FALSE, fig.cap="Input parameter values interactively for parameterized reports.", out.width="100%"}
knitr::include_graphics("images/params-input.png", dpi = NA)
```
The input controls for different types of parameters can be customized by specifying additional sub-items within the parameter specification in YAML. For example, sliders, check boxes, and text input boxes can all be used for input controls.
In addition, we can also specify constraints of the values allowed in each parameter. For example, we may only want our model to be run for years between 2010 and 2018. This is particularly beneficial if you would like other users to interact with the report, as it prevents users from attempting to run reports outside of the designed limits.
Adapting our above example to include some settings:
```yaml
---
title: My Document
output: html_document
params:
year:
label: "Year"
value: 2017
input: slider
min: 2010
max: 2018
step: 1
sep: ""
region:
label: "Region:"
value: Europe
input: select
choices: [North America, Europe, Asia, Africa]
printcode:
label: "Display Code:"
value: TRUE
data:
label: "Input dataset:"
value: results.csv
input: file
---
```
This results in the user interface for the parameters as shown in Figure \@ref(fig:params-controls).
```{r params-controls, echo=FALSE, fig.cap="Custom controls for parameters.", out.width="100%"}
knitr::include_graphics("images/params-controls.png", dpi = NA)
```
The type of Shiny control used is controlled by the `input` field. Table \@ref(tab:params-types) shows the input types currently supported (see the help page for the associated Shiny function for additional attributes that can be specified to customize the input, e.g., `?shiny::checkboxInput`).
```{r, params-types, echo=FALSE}
knitr::kable(
read.table(
text = "checkbox checkboxInput
numeric numericInput
slider sliderInput
date dateInput
text textInput
file fileInput
radio radioButtons
select selectInput
password passwordInput",
col.names = c("Input Type", "Shiny Function"),
check.names = FALSE
),
caption = "Possible input types and the associated Shiny functions for parameterized reports.",
booktabs = TRUE
)
```
## Publishing {#params-publish}
Parameterized reports are supported by the publishing platform RStudio Connect (https://www.rstudio.com/products/connect/). If you publish a parameterized report to an RStudio Connect server, you will be able to compile reports by interactively choosing different parameter values on the server, and easily store/navigate through different reports built previously. You may watch a video demonstration at https://bit.ly/rsc-params.