# get dimensions
dim(reviews_fr)

## [1] 20508    15

# get variable names
names(reviews_fr)

##  [1] "points"                "title"                
##  [3] "description"           "taster_name"          
##  [5] "taster_twitter_handle" "price"                
##  [7] "designation"           "variety"              
##  [9] "region_1"              "region_2"             
## [11] "province"              "country"              
## [13] "winery"                "region_type"          
## [15] "vintage"

Single Plots

To begin, I will create a basic scatterplot concerning the relationship between the points a wine receives and the year in which it was produced. The syntax structure is similar to that of ggplot2: one initially creates an empty figure and then adds layers. You can use the pipe operator between layers (rather than the plus sign of ggplot2).

# load package
library(rbokeh)

# points versus vintage
figure() %>% 
  ly_points(x = vintage,
            y = points,
            data = reviews_fr)

This very simple scatter plot is not very informative. It is difficult to see much of a trend, likely for two reasons: (1) the actual trend—or lack thereof—in the data and (2) the size, density, and transparency of the points.

Fixing the second set of issues is relatively straightforward. I will set the point size to 7 and the alpha to 0.3.

# adjust size and alpha
figure() %>% 
  ly_points(x = vintage,
            y = points,
            data = reviews_fr,
            size = 7,
            alpha = 0.3)

The darker points indicate considerable overlap of observations, and the smaller dots allow us to see more precisely where these denser regions are located, which are unfortunately rather frequent. This issue lies in the data, so I will ignore it for now because the goal is to showcase rbokeh.

The scatterplot currently has no title, and the axis labels are not capitalized, as they are simply the variable names. In order to fix this, we specify the title, xlab, and ylab options.

figure(title = "Rating versus vintage of French wines") %>% 
  ly_points(x = vintage,
            y = points,
            data = reviews_fr,
            size = 7,
            alpha = 0.3,
            xlab = "Vintage",
            ylab = "Points")

This version looks better, but the axis labels are larger than the plot title. We can fix this issue with the theme_title() function.

figure(title = "Rating versus vintage of French wines") %>% 
  ly_points(x = vintage,
            y = points,
            data = reviews_fr,
            size = 7,
            alpha = 0.3,
            xlab = "Vintage",
            ylab = "Points") %>% 
  theme_title(text_font_size = "13pt")

Et voilà!

My favorite capability of the robkeh visualizations is having information on each observation presented to you upon hovering above the point. To make this occur, we use the hover option.

figure(title = "Rating versus vintage of French wines") %>% 
  ly_points(x = vintage,
            y = points,
            data = reviews_fr,
            size = 7,
            alpha = 0.3,
            xlab = "Vintage",
            ylab = "Points",
            hover = c(vintage, points)) %>% 
  theme_title(text_font_size = "13pt")

One issue with the information provided when one hovers over a point is that the label for the values is simply a variable name. There are two different ways to fix this problem. One may reference within a string the variable preceded by an @ symbol (e.g., “@vintage”), which enables the possibility for hovers to include something like “This wine was produced in @vintage”, but, in my experience until now, you cannot include multiple pieces of information. The other way to cutomize the hover is to use html code. While this might seem a bit clunky at first, you will additionally have the ability to include multiple pieces of information, and the presentation is more flexible than the large table the accompanies the previous example. I will illustrate now the use of html to alter the hover information.

# customize hover
figure(title = "Rating versus vintage of French wines") %>% 
  ly_points(x = vintage,
            y = points,
            data = reviews_fr,
            size = 7,
            alpha = 0.3,
            xlab = "Vintage",
            ylab = "Points",
            hover = "<b>Vintage</b>: @vintage <br>
            <b>Points</b>: @points") %>% 
  theme_title(text_font_size = "13pt")

Multiple Plots

Unlike ggplot2, rbokeh does not have a succinct way to create multiple plots (e.g., a facet option). Instead, you have to create the plots you want to be grouped together and then use the grid_plot() function to join them together. Once you do this, customizing the group is relatively straightforward. I switch in this section to visualizing the relationship between price and rating for a few wine regions within France.

The first task consists of splitting data by region, which the split() function makes easy.

# split by region/province
split_province <- split(reviews_fr, reviews_fr$province)

Here are the regions found in the data.

unique(reviews_fr$province)

##  [1] "alsace"               "beaujolais"           "bordeaux"            
##  [4] "burgundy"             "france other"         "southwest france"    
##  [7] "rhône valley"         "languedoc roussillon" "provence"            
## [10] "loire valley"         "champagne"

Next, I write a function to create a plot for each function that I will then apply to all of the sub-sets of region data. Within each plot I include the price and points values in the hover.

# create plotting function
plot_scatter <- function(x){
  figure() %>% 
    ly_points(x = price,
              y = points,
              data = x,
              hover = "<b>Price</b>: @price<br><b>Points</b>: @points",
              size = 6,
              alpha = 0.3)
}

# create list of figures
scatter_list_province <- lapply(split_province, plot_scatter)

I do not want plots for all 11 regions found within the data, so I will choose four regions: Bordeaux, Burgundy (Bourgogne), Champagne, and Alsace.

# now plot together
grid_plot(list(Bordeaux = scatter_list_province[["bordeaux"]],
               Burgundy = scatter_list_province[["burgundy"]],
               Champagne = scatter_list_province[["champagne"]],
               Alsace = scatter_list_province[["alsace"]]),
          nrow = 2,
          same_axes = TRUE,
          xlim = c(0, 1000))

It is important in the initial list provided to type on the right hand of the equal sign the plot name as you would like it to appear; I have not (yet, hopefully) found away to easily alter the titles in another manner, for example, if one wants multiple words. I specified two rows to ensure that I received a 2 by 2 table of figures. rbokeh also allows you to specify the number of columns. The same_axes argument ensures that all plots have the same ranges, and I set the limit of the x axis as 1000. I chose to do so because there are extreme values in some of the plots, which made trends even more difficult to see easily. Note however that these points are still on the plot: because the scatter plots are interactive, you may click and drag the plot past the 1000 mark on the x axis.

Using Color within a Single Plot

In this final section, I return to creating a single visualization and will examine the wines from Alsace. I begin by subsetting to reviews_fr data to include only wines from this region.

# filter alsace
alsace <- reviews_fr %>% 
  filter(province == "alsace")

I will include in this figure all of the previous customization types I have covered until now. I provide a title with 14 point font, a hover that include information on the vintage, price, and points, resized and largely transparent dots, and x and y axis labels. The key change is that I add a color option, using taster_name as a factor variable (it is currently a character string). I also include height and width specifications, but these simply enlarge the resulting figure. There are

figure(width = 600,
       height = 600,
       title = "Alsace Wine Reviews, 1996-2016") %>% 
  ly_points(x = price,
            y = points,
            data = alsace,
            hover = "<b>Winery</b>: @winery<br>
            <b>Vintage</b>: @vintage<br>
            <b>Price</b>: $@price<br>
            <b>Points</b>: @points<br>",
            size = 6,
            alpha = 0.3,
            color = as.factor(taster_name),
            xlab = "Price",
            ylab = "Points") %>% 
  theme_title(text_font_size = "14pt")

In addition to the bivariate distribution of scores (points) againt prices, we also see a general trend in the reviewing for Alsace. Ann Krebiehl (blue) tends to review wines with higher scores than Roger Voss (red). Joe Czerwinski reviewed only three wines, and Lauren Buzzeo reviewed two. Hovering over each observation provides information on the winery, vintage, price, and points a wine received.

It is important to note three issues with the figure. First, the dollar sign is not always placed consistently in the hover. If the price is three digits, then there is no space between the sign and the first digit, but, if the price is two digits, then there is a space. I have not yet found a way remedy this problem. Second, the legend is in the way. I will move it—using the legend_location option in the figrue() function—to the bottom right of the figure in the next iteration. Finally, the variable name, which appears in the legend, is exactly as I coded it when creating the figure; this means that it includes as.factor() and has an underscore rather than a space between the two words. The as.factor() problem may be addressed by changing the variable’s class before plotting the data, but I have not found a way to customize the presentation of the variable name.

# recode variable
alsace <- alsace %>% 
  mutate(Taster_Name = as.factor(taster_name))

# plot
figure(legend_location = "bottom_right",
       width = 600,
       height = 600,
       title = "Alsace Wine Reviews, 1996-2016") %>% 
  ly_points(x = price,
            y = points,
            data = alsace,
            hover = "<b>Winery</b>: @winery<br>
            <b>Vintage</b>: @vintage<br>
            <b>Price</b>: $@price<br>
            <b>Points</b>: @points<br>",
            size = 6,
            alpha = 0.3,
            color = Taster_Name,
            xlab = "Price",
            ylab = "Points") %>% 
  theme_title(text_font_size = "14pt")

Note the Toolbar

The toolbar is an important figure of rbokeh plots. You can find it in the above figure on the upper right side, and it is composed of several semi-transparent icons: pan, box zoom, wheel zoom, reset, a selector to enable or disable the hover tool, and a link to more information about rbokeh. There are many ways to customize the toolbar, including removing it altogether. You change its location by including “left”, “right”, “top”, “bottom” after the toolbar_location option in figure(); if you put NULL rather than one of the strings, it removes the toolbar.

I particularly enjoy the zoom features of rbokeh. Having over 20,000 observations, as does the reveiws_fr data, means that certain portions of the scatter plot can get quite dense, and zooming lets you inspect a smaller portion of the data while retaining the ability to examine other spaces on the plot. The box zoom enables you to enlarge a set of points that you select (i.e., that fall within a box that appears when you click and drag the cursor across the figuree). The wheel zoom allows you to use you mouse’s scroll wheel (or function if you have a touch mouse) to zoom in and out on portions of the plot. If you decide you no longer want to zoom, reset will make the plot appear as it was originally coded.[^2]

You can of course change the default presentation from which the toolbar enables you to deviate. I provide below an example in which I include only pan, box zoom, and reset as options; this is accomplished with the tool option in figure(). You will still have the option to disable to hover tool.

# change tools
figure(legend_location = "bottom_right",
       width = 600,
       height = 600,
       title = "Alsace Wine Reviews, 1996-2016",
       tools = c("pan", "box_zoom", "reset")) %>% 
  ly_points(x = price,
            y = points,
            data = alsace,
            hover = "<b>Winery</b>: @winery<br>
            <b>Vintage</b>: @vintage<br>
            <b>Price</b>: $@price<br>
            <b>Points</b>: @points<br>",
            size = 6,
            alpha = 0.3,
            color = Taster_Name,
            xlab = "Price",
            ylab = "Points") %>% 
  theme_title(text_font_size = "14pt")