plot | A repository for plotting and visualizing data | Data Visualization library

Found the problem.
I was installing pandas_profiling, and this package updated pyyaml to version 6.0 which is not compatible with the current way Google Colab imports packages.
So just reverting back to pyyaml version 5.4.1 solved the problem.

For more information check versions of pyyaml here.
See this issue and formal answers in GitHub

##################################################################
For reverting back to pyyaml version 5.4.1 in your code, add the next line at the end of your packages installations:

Looking at the plotly code here (excerpt below), it seems that the raster is only defined for any values of x and y available in the dataset - and whatever happens in between is up the the rest of the plotly code.

The long execution time comes from the fact that each child of a Parent registers a listener with the disabled and treeVisible properties of that Parent. The way JavaFX is currently implemented, these listeners are stored in an array (i.e. a list structure). Adding the listeners is relatively low cost because the new listener is simply inserted at the end of the array, with an occasional resize of the array. However, when you remove a child from its Parent and the listeners are removed, the array needs to be linearly searched so that the correct listener is found and removed. This happens for each removed child individually.

So, when you clear the children list of the Group you are triggering 1,000,000 linear searches for both properties, resulting in a total of 2,000,000 linear searches. And to make things worse, the listener to be removed is either--depending on the order the children are removed--always at the end of the array, in which case there's 2,000,000 worst case linear searches, or always at the start of the array, in which case there's 2,000,000 best case linear searches, but where each individual removal results in all remaining elements having to be shifted over by one.

There are at least two solutions/workarounds:

1. Don't display 1,000,000 nodes. If you can, try to only display nodes for the data that can actually be seen by the user. For example, the virtualized controls such as ListView and TableView only display about 1-20 cells at any given time.

2. Don't clear the children of the Group. Instead, just replace the old Group with a new Group. If needed, you can prepare the new Group in a background thread.

   Doing it that way, it took 3.5 seconds on my computer to create another Group with 1,000,000 children and then replace the old Group with the new Group. However, there was still a bit of a lag spike due to all the new nodes that needed to be rendered at once.

   If you don't need to populate the new Group then you don't even need a thread. In that case, the swap took about 0.27 seconds on my computer.

In addition to the provided comments:

%$% also called the Exposition pipe vs. %>%:

This is a short summary of this article https://towardsdatascience.com/3-lesser-known-pipe-operators-in-tidyverse-111d3411803a

"The key difference in using %$% or %>% lies in the type of arguments of used functions."

One advantage, and as far as I can understand it, for me the only one to use %$% over %>% is the fact that we can avoid repetitive input of the dataframe name in functions that have no data as an argument.

For example the lm() has a data argument. In this case we can use both %>% and %$% interchangeable.

But in functions like the cor() which has no data argument:

This is a good question... and I would have hoped there would be a practical guide somewhere. One could question if SO would be a good place to ask this question, but regardless, here's my attempt to summarize the various scale_color_*() and scale_fill_*() functions built into ggplot2. Here, we'll describe the range of functions using scale_color_*(); however, the same general rules will apply for scale_fill_*() functions.

There are 22 functions in all, but happily we can group them intelligently based on practical usage scenarios. There are three key criteria that can be used to define practically how to use each of the scale_color_*() functions:

1. Nature of the mapping data. Is the data mapped to the color aesthetic discrete or continuous? CONTINUOUS data is something that can be explained via real numbers: time, temperature, lengths - these are all continuous because even if your observations are 1 and 2, there can exist something that would have a theoretical value of 1.5. DISCRETE data is just the opposite: you cannot express this data via real numbers. Take, for example, if your observations were: "Model A" and "Model B". There is no obvious way to express something in-between those two. As such, you can only represent these as single colors or numbers.

2. The Colorspace. The color palette used to draw onto the plot. By default, ggplot2 uses (I believe) a color palette based on evenly-spaced hue values. There are other functions built into the library that use either Brewer palettes or Viridis colorspaces.

3. The level of Specification. Generally, once you have defined if the scale function is continuous and in what colorspace, you have variation on the level of control or specification the user will need or can specify. A good example of this is the functions: *_continuous(), *_gradient(), *_gradient2(), and *_gradientn().

We can start off with continuous scales. These functions are all used when applied to observations that are continuous variables (see above). The functions here can further be defined if they are either binned or not binned. "Binning" is just a way of grouping ranges of a continuous variable to all be assigned to a particular color. You'll notice the effect of "binning" is to change the legend keys from a "colorbar" to a "steps" legend.

The continuous example (colorbar legend):

Great question. I have often thought about this. I don't know of any packages that allow it natively, but it's not terribly difficult to do it yourself, since geom_text accepts angle as an aesthetic mapping.

Say we have the following plot:

I think you might be slightly overcomplicating things. Ideally, you'd just want a single key drawing method for the whole layer. However, because you're using a Stat to do the majority of calculations, this becomes hairy to implement. In my answer, I'm avoiding this.

Let's say I'd want to use a geom-only implementation of such a layer. I can make the following (simplified) class/constructor pair. Below, I haven't bothered width_scale or height_scale parameters, just for simplicity.

This is an interesting problem. The short answer: Yes, it's possible. But I don't see a way around hard coding the position of table and legend, which is ugly.

The suggestion below requires hard coding in three places. I am using {ggpubr} for the table, and {cowplot} for the stitching.

Another problem arises from the legend key spacing for vertical legends. This is still a rather unresolved issue for other keys than polygons, to my knowledge. The associated GitHub issue is closed The legend spacing is not a problem any more. Ask teunbrand, and he knows the answer.

Some other relevant comments in the code.

You can do calculations within a function for the x and y values to construct the ggplot which extends the circle all the way round and gives labels correct heights.

I've adapted a function to work with other datasets. This takes a dataset in a tidy format, with:

• a 'year' column
• one row per 'event'
• a grouping variable (such as country)

I've used Nobel laurate data from here as an example dataset to show the function in practice. Data setup: