Dplyr 0.7.0 has been published. One of the greatest improvements is the enhancement for standard evaluation.

Let’s look at an example. Let’s say I want to apply a function to a column of a data.frame. But the name of the data.frame can change from call to call. So the actual column is the value of a string.

Earlier to dplyr 0.7.0 I used the following methods using lazyeval:

One of the most annoying features of ggplot2 was the missing x-axis scale on facet-plots with some shorter columns.

To illustrate the problem see the following (pre-2.2-version) example:

I recently used the function spread with a factor. Spread is defined in Hadley Wickham‘s package tidyR. It’s used to convert a data.frame from long format to wide format.

Here’s an example showing my problem:

Let’s say you have some data like this 1 2 3 4 5 set.seed(42) data <- data.frame( x = seq(1,30), y = 0.3 * seq(1,30) + 2 * rnorm(30) ) So let’s plot it using ggplot: 1 2 library(ggplot2) ggplot(data, aes(x=x, y=y)) + geom_point() + geom_line() Plotting an aim-corridor If this is something like your revenue, pageviews or something like that you may ask: “Is this good or is this bad?

Let’s say you have to do the same analysis with diffent data and you want to use knitr’s RMarkdown to publish the results of each analysis in one report. The Subview or Parameterized Report So you start writing a RMarkdown-file. You can define several variables this RMarkdown-files depends on. It’s called Parameterized Reports. As the above article describes you define these parameters in the header of the RMarkdown-file. So let’s try it: