borisdelange / linkr-content Goto Github PK

Go to the "Scripts management" tab to add, delete, or rename scripts.

A script is a text file containing code.

Once a script is created, you can edit it in the "Script" tab and execute the code.

Write R code and execute it: the result returned by the console will appear below the text editor.

To know the data model used by the application, click on the help button (question mark at the top right of the page) when you are on the "Data / Access Data" page.

Example code (get the min, max, and average heart rate of our patients):

d$measurement %>%
    dplyr::filter(measurement_concept_id == 3027018) %>%
    dplyr::group_by(person_id) %>%
    dplyr::summarize(min_weight = min(value_as_number), max_weight = max(value_as_number), avg_weight = mean(value_as_number)) %>%
    dplyr::ungroup()

You can also write code in Rmarkdown.

This is Markdown to which you can add R code.

For more information on RMarkdown, visit their site.

Example code (the same example as above, in RMarkdown - a backslash has been added to prevent code execution):

# Script on heart rate

\```{r}
d$measurement %>%
    dplyr::filter(measurement_concept_id == 3027018) %>%
    dplyr::group_by(person_id) %>%
    dplyr::summarize(min_weight = min(value_as_number), max_weight = max(value_as_number), avg_weight = mean(value_as_number)) %>%
    dplyr::ungroup()
\```

You can create figures, for example with the ggplot2 library.

Here's an example of code:

# A list containing the data for the plot
data <- list()

# Filter data
data$x <- d$measurement %>% dplyr::filter(measurement_concept_id == 3027018)

# Create ggplot2 plot
data$x %>%
    ggplot2::ggplot(ggplot2::aes(x = value_as_number)) +
    ggplot2::geom_histogram(bins = 50, fill = "#377EB8", color = "#FFFFFF") +
    ggplot2::theme_minimal() +
    ggplot2::labs(x = "Heart rate (bpm)", y = "")

In order to save multiple plots to a single widget, you can create scripts from the "Scripts management" tab.

You can also rename and delete scripts from this tab.

A script includes:

• The parameters saved on a plot, from the "Plot" tab.
• The code corresponding to that plot, from the "Code" tab.

Once a script is created, select it from the "Plot" or "Code" tab.

Save this script after making changes.

Go to the "Plot" tab.

Choose the variable to display on each axis in the "Variables" tab.

Some plots will have a variable on only one axis (geom_histogram), while others will only display if variables are specified on both axes (geom_point).

For each plot, choose:

• The theme of the plot.
• The text for the x axis.
• The text for the y axis (whether a variable is assigned to it or not).

In the "Variables" tab, it's possible to group data, by patient or by time. Then choose the function to apply to group the data.

For example:

• Group data by patient by selecting the "maximum" function: the maximum value of the parameter, across all stays, will be chosen for each patient.
• Group data every 4 hours by selecting the "average" function: the maximum value will be chosen for each 4-hour interval, for each patient (thus, multiple data points per patient).

By checking "Show plot at script launch", the plot will be displayed upon the script's loading.

By clicking on "Generate Code" from the "Plot" tab, the code corresponding to the plot will be generated in the text editor, in the "Code" tab.

You can edit this code and save your changes.

If you click on Generate Code again, your code will be deleted and replaced by the default code corresponding to the plot configured in the "Plot" tab.

Click on "Run code" to display the plot corresponding to the code.

By checking "Run code at script launch", the code will be executed upon the script's loading.

Allows visualization of the distribution of a single continuous variable by dividing the x-axis into intervals and counting the number of observations in each interval.

You can choose the size of the bars in two ways:

• Either by the size of the bars, depending on the x-axis (a size of 50 means a bar will take up 50 of the unit specified on the x-axis).
• Or by the number of total bars displayed.

Allows visualization of the distribution of two continuous variables, one on the x-axis and the other on the y-axis.

By default, only points having a value at a given moment both for the x variable and for the y variable will be displayed.

For instance, if I choose systolic blood pressure for x and mean arterial pressure for y (as in the example), and I have an x value on a certain day at 18:37:10 and a y value on the same day at 18:37:11, since the moments aren't exactly the same, there won't be a corresponding point on the graph.

To counter this, it is possible to group data, by patient or by time (see Usage section).

Go to the "Scripts management" tab to add, delete, or rename scripts.

A script is a text file containing code.

Once a script is created, you can edit it in the "Script" tab and execute the code.

Write R code and execute it: the result returned by the console will appear below the text editor.

To know the data model used by the application, click on the help button (question mark at the top right of the page) when you are on the "Data / Access Data" page.

Example code (get the min, max, and average heart rate of our patients):

d$measurement %>%
    dplyr::filter(measurement_concept_id == 3027018) %>%
    dplyr::group_by(person_id) %>%
    dplyr::summarize(min_weight = min(value_as_number), max_weight = max(value_as_number), avg_weight = mean(value_as_number)) %>%
    dplyr::ungroup()

You can also write code in Rmarkdown.

This is Markdown to which you can add R code.

For more information on RMarkdown, visit their site.

Example code (the same example as above, in RMarkdown - a backslash has been added to prevent code execution):

# Script on heart rate

\```{r}
d$measurement %>%
    dplyr::filter(measurement_concept_id == 3027018) %>%
    dplyr::group_by(person_id) %>%
    dplyr::summarize(min_weight = min(value_as_number), max_weight = max(value_as_number), avg_weight = mean(value_as_number)) %>%
    dplyr::ungroup()
\```

You can create figures, for example with the ggplot2 library.

Here's an example of code:

# A list containing the data for the plot
data <- list()

# Filter data
data$x <- d$measurement %>% dplyr::filter(measurement_concept_id == 3027018)

# Create ggplot2 plot
data$x %>%
    ggplot2::ggplot(ggplot2::aes(x = value_as_number)) +
    ggplot2::geom_histogram(bins = 50, fill = "#377EB8", color = "#FFFFFF") +
    ggplot2::theme_minimal() +
    ggplot2::labs(x = "Heart rate (bpm)", y = "")