The 'Add sample means' function uses a '+' sign to indicate the mean. This is OK when showing a box plot (given the median is indicated by a bar) but looks weird when showing a dot plot (e.g. for small sample sizes). Would be great if the mean was indicated by a bar instead, or if there was an option to choose either a plus sign or a bar to indicate the mean.

Assuming there 5 same boxplots created in a graph, the data point of the first boxplot with jitter option will still have over-plotting problem. It can be seen that the over plotting problems decrease from one boxplot to another from left to right.

Notify user when bean plot 'decides' to plot data on log-scale.

It would be great to be able to save all settings for a boxplot (separator, titles, axis labels etc.) so that one can easily go back and make changes.

Since the control panel can get quite 'long', it would be useful to 'uncouple' the control panel and the plot area so that the plot is always visible.

If median is, for example, '0.002', the statistics table shows '0'.

If numbers on y-axis are big (many digits), they might overlap with y-axis label.

When displaying data with boxplots, add an option to not show the outliers ("outline=F")

Use case: If there are strong outliers, a user might want to focus on the boxplots and not show any outliers.

Implement option to add values of boxplot statistics directly to the plot (median, quartiles). Allow single data points to be shown - not sure how to identify the data point (row number?) and if data points in row number x should be shown in all boxplots if data set has several columns.

Longer labels are abruptly cut off toward the left end of the label. Would be great if this was not the case.

Hello @VizWizard thanks for this beautiful software.
I am trying to add new datapoints to my thesis and tried to visualise via the website but whatever I do, I cannot seem to get any visualisations.

I have tried using the "load sample data" option, with default settings to no avail.
I used the website to obtain visualisations as late as third quarter of last year.
I've also tried using the exact same datasets I used last year, but also to no avail.

I've checked through the commit logs and notice some new commits some 18 days ago.

I've done some poking around on the website and see no console errors or even API calls being made.

I would try to follow the instructions on running the project locally, and try to point to earlier commits to see if that solves my issue. Thank you.

If there are more than 500 data points, only 100 rows are shown in the table which can be confusing.

When data points are added to the plot (default, bee swarm or jitter), outliers appear twice because they are plotted with the original boxplot and the again when the data points are added.
Solution: When points are to be added, draw box plots with outline=FALSE when adding data points and set ylim = c(min(data), max(data)). Keep in mind orientation of boxplot (ylim vs xlim).

In medicine, the 2.5th and 97.5th percentiles are frequently used - this is called reference range or prediction interval.

beanplot --> https://cran.r-project.org/web/packages/beanplot/index.html
vioplot --> https://cran.r-project.org/web/packages/vioplot/index.html
beeswarm -> http://www.cbs.dtu.dk/~eklund/beeswarm/

It might be useful to display row numbers for outliers to identify outliers in data set.

Change implementation so that special characters (spaces, dashes etc.) are not changed to '.'.