We can run an SQL query with the dbGetQuery function which expects a connection object and a string with the SQL commands. If we wanted to take a subset of the CO2 dataset and view only those observations taken in Quebec we could use the following code:

> dbGetQuery(con, paste("select * from co2 where Type = 'Quebec'"))
   row_names Plant   Type  Treatment conc uptake
1          1   Qn1 Quebec nonchilled   95   16.0
2          2   Qn1 Quebec nonchilled  175   30.4
...
41        41   Qc3 Quebec    chilled  675   39.6
42        42   Qc3 Quebec    chilled 1000   41.4

We can narrow the search by specifying a condition on one of the other columns, for example limit the search to concentrations under 300 units:

> dbGetQuery(con, paste("select * from co2 where Type = 'Quebec' and Conc < 300"))
   row_names Plant   Type  Treatment conc uptake
1          1   Qn1 Quebec nonchilled   95   16.0
2          2   Qn1 Quebec nonchilled  175   30.4
...
17        37   Qc3 Quebec    chilled  175   21.0
18        38   Qc3 Quebec    chilled  250   38.1

As a final example we could look for all observations with a concentration less than 300 units buy order the data by increasing values of uptake:

> dbGetQuery(con, paste("select * from co2 where Conc < 300 order by uptake"))
   row_names Plant        Type  Treatment conc uptake
1         71   Mc2 Mississippi    chilled   95    7.7
2         29   Qc2      Quebec    chilled   95    9.3
...
35        38   Qc3      Quebec    chilled  250   38.1
36        17   Qn3      Quebec nonchilled  250   40.3

This only scratches the surface of the SQL statements that could be run to extract data from a MySQL database.

MySQL is a popular database system and there is a library RMySQL that can be used to access this database – it is important to ensure that the version of MySQL matches with the R library. If there isn’t a match then the system might exhibit erratic behaviour.

The first step is to make the RMySQL library available in the working session:

library(RMySQL)

If this command runs without problems then we need to create a connection object for our session:

con = dbConnect(MySQL(), dbname = "test", user = "user1",
  password = "pwd1")

The first argument to this function creates and initializes a MySQL client. It returns an driver object that allows you to connect to one or several MySQL servers. The dbname argument is used to specify the name of the database and the user and password arguments should be self-explanatory.

After creating our connection successfully we can get R to list the tables that are stored in this database:

> dbListTables(con)
[1] "co2"

In this example there is only one table and its name is returned by the dbListTables function. To read the data from this table we use the dbReadTable function and specify the connection object as well as the name of the table in the MySQL database:

> dbReadTable(con, "co2")
   Plant        Type  Treatment conc uptake
1    Qn1      Quebec nonchilled   95   16.0
2    Qn1      Quebec nonchilled  175   30.4
3    Qn1      Quebec nonchilled  250   34.8
...

We can save the table to a data frame object rather than the default action of printing to the console.

After undertaken some analysis we might want to save a data set to the database and the dbWriteTable function is used:

> dbWriteTable(con, "CO2", data.frame(CO2),
  overwrite = TRUE)
[1] TRUE

The first argument is the connection object, the second is the name that the table will be referred to in the database and the third argument is the data to be saved. In this case we have used the overwrite argument to copy over any existing table of the same name.

We can delete a table using the dbRemoveTable function:

> dbRemoveTable(con, "CO2")
[1] TRUE

and when we reach the end of our need for the connection, the dbDisconnect function will remove the connection that we have been using:

> dbDisconnect(con)
[1] TRUE