crosstalk

I use crosstalk in order to be able to click on a bar in the barplot and subset the table below based on the Abcd. For some reason if I press on a bar I get only one row of the dataset while I should get all of them. If I select all the bars again the table still displayes only one row.

I have the shiny app below in which I use crosstalk package to create interaction between the chart and the table. I would like to ask if it is possible to select more than one bars at the same time in order to bring back the table to its initial form and also how can you unselect a bar without having to click on another one?

I am trying to change a plotly graph via crosstalk in my Rmarkdown doc. This works in principle fine, however, I want to set the dropdown to the first element on startup via JavaScript. This change, however, does not update the graph. If I run the same code once the full page is rendered, or use a delay it works.

Thus, I assume that at the point the update JavaScript is run, the plotly graph is not yet rendered. The setTimeout solution works, but this feels hackis. So what is the canonical way to inform plotly/crosstalk that the input changed?

In the gif you see that upon startup the plot shows the full range, only after changing to b and then to a updated the graph accordingly. I would like that the graph is correctly filtered already upon startup.

I am trying to modify the appearance of a crosstalk filter slider by changing its colour and font. There is no built-in option to do this within the filter_slider() function, so I looked up the code behind the function to see if it specifies the colour and font of the output. I found nothing that indicates that it does, so I was wondering if it is possible to add some lines to the function that enable changing the colour of the slider and its font. I have very limited knowledge of writing functions, so I do not know how to modify a complicated function like this one. I am attaching the code behind the filter_slider() function below.

I make a leaflet map with a responsive heatmap using addHeatmap. Unfortunately, this kind of tool it is not enough useful because two main problems: 1) The heatmap is redrawed with each new level of zoom and 2) you can not make the heatmap and the points in a separated group each one.

It is possible a similar solution with addWebGLHeatmap?

There is the code for the addHeatmap solution, following this question

In an Rmarkdown html document, how does one select a default value for a crosstalk::filter_select dropdown that will work with plotly plots? E.g., in the example below, to have just group 'a' selected when the RMD is knitted.

I know that for the reprex below example using plotly buttons would be easier, but when there are more than 4-5 or so choices the plotly dropdowns/buttons take up too much room/are quite ugly. Also hoping to avoid running a shiny server, the idea is to have everything running client side for speed purposes.

There is a PR in crosstalk that adds a "default choice" argument to the filter_select function, but that version doesn't work with plotly (https://github.com/rstudio/crosstalk/pull/70). I would guess the easiest way would be to add javascript to the doc to manipulate the crosstalk object, but a few experiments haven't gotten very far yet.

Reprex rmd:

I want to make a map with points and a responsive heatmap using crosstalk in R. Like this:

I'm new to HTML and interactive tables, and I'm having trouble building a table container to add headers for my data set. Is there an easy way to insert table headers for my data set? I want columns 2-5 (excluding date in column 1) and columns 6-9 to have header 'Sector' and 'Industry' respectively.

I've included an extract of the dataset below.

Thank you!

I'm trying to create a map visualization in plotly using mapbox (plotly::plot_mapbox()), where the points in the map belong to different geographical districts ("comunas") and the user can choose a geographical district in a drop-down widget created by crosstalk::filter_select().

The expected behavior is that, when a geographical district is selected in the widget, it should trigger a filter in the map, which should only show the points that belong to the selected district.

However, I'm unable to make it work (I'm following the instructions presented here). The code that I wrote successfully creates 1) the map and 2) the filtering widget, but the components don't interact/talk to each other, i.e. when I select a district in the drop-down, there is no filtering effect in the map, it keeps showing all the points, not only those who belong to the selected district.

Am I doing something wrong? How can I make this work as intended?

Below is a reproducible example with a subset of the original data (it requires a Mapbox token to run):