cla | Source Implementation of the Critical-Line Algorithm | Portfolio library

Based on the comment by Jody Klymak I got what I wanted with the cairo backend even without snap=False, see attachement.

I'm sure the intention is not to enter actual values into your code as you show in comments.

Regarding the loop(s) arrangement, consider to read one value from table A, then check that value against every value in table B. Then again read next value from A, and check again against all values in B and so on... This requires that the loops are nested

Your problem is that you defined the subplot using matplotlib.pyplot (plt.).

self.area_plot.figure, self.ax = plt.subplots() <-- here

If you register the subplot on your defined figure (and clean the code a bit up), it should work as expected:

The reason is the Circle patch is drawn as a face and an edge - the edge has a thickness, which is what is overlapping.

We can resolve this by setting the facecolor to red, and the edgecolor to None, so it doesn't get plotted.

For example:

This is caused by the recently released matplotlib 3.5. We are planning on releasing a fix for this with MetPy 1.2, scheduled to be released shortly.

In the meanwhile as a workaround, you'll need to downgrade to matplotlib 3.4.

Problem is not with plotly package but with the way you have created the data.frame dat.

Also one correction when the code is working that time you have considered five distinct years and repeated the years 4 times whereas when the code isn't working as expected you have considered four distinct years and repeated five times. So when warning is occurring both cla A and B have no common year between them.

cla A has 2018 and 2020 whereas cla B has 2019 and 2021 as the distinct years which is causing the warning. This is happening because of rep(c(2018:2021), 5) which creates the data frame as

Answering your last question first, your big issue is that you are trying to use one figure to display two seperate graphs, without understanding how pyplot actually works. It is possible to do so, as I will outline below, but you lose any modifications you make to the graphs when you switch between them. I would recommend reading the working with multiple axes and figures section of the pyplot tutorial. The key takeaway as that in MATLAB and pyplot, all plotting functions are applied to the last figure and axes.

When you call self.draw_graph_two(), the last axes you create are the RadioButtons. Therefore, pyplot has these axes as the current axes. When you then call self.draw_graph_one() for the second time (the first time was when you intialised), it draws the plot on the current figure and current axes, where the RadioButtons are.

To solve this, call plt.clf() as the first line of draw_graph_one and draw_graph_two, before attempting to draw the new graph. This clears the figure, allowing for a new plot to be drawn from scratch (which is what your two draw functions do). This allows you to continuously switch between your two plots, although each time you switch the plot is cleared and any modifications to the plot are lost. You can also remove the self.ax.remove() lines, as they cause errors and are not needed.

For your first two questions, these can be answered by reading the documentation for the RadioButtons and Slider here. You must keep a reference to the widgets for them to remain responsive. So modifying your draw_graph_two to return alfa_slider and radio keeps these widgets responsive. Alternatively, declaring them as instance variables also keeps a reference (self.alfa_slider and self.radio).

So your working graph viewer is:

1. You already created a Figure object and an ax, so always plot on it and update your canvas.
2. Use tkinter.after method to call your function recursively so it won't freeze your GUI.

Below are the minimum changes required to showcase how it can be done:

User code cannot run on my platform, so just example here. Of course, you can use sg.Canvas here.