@story645 thanks for the suggestion! This PR is ready for review

It looks like something went wrong with the layout. Also, please read the comments on the issue for ideas how to improve this.

thanks for the feedback, I tried to make some adjustments but I cannot see if my changes were successful, how does this link you commented before work?

how does this link you commented before work?

At the bottom of this page are some automated checks. Scroll to the bottom of them and you see "View the built docs". Click on that and you can then navigate to your example from there.

But you can also just run your script locally to see how the plot looks. That will give you faster feedback.

Thanks a lot for all the suggestions, I tried to see the changes locally like@rcomer said and it should look much better now, please let me know if there's still something missing

Can you please try the dataset@jklymak suggested in the original issue?

Also can you remove the spines (code should also be in the original example). Also please remove the yaxis_inversion as that's confusing me here - it might be clearer if you add the colorbar like in the original issue. Which also setting the margins to 0 (see original example) will remove the spacing for the bar chart so that it takes up the whole space (also please set height to 100% so they line up) - you are welcome to add that as a comment if it makes it clearer and to ask about the purpose of any code you don't understand.

I tried the first example from here :https://matplotlib.org/devdocs/gallery/images_contours_and_fields/image_demo.html and it worked just fine! please tell me if it is alright now

The new dataset isn't a normal distribution so the histogram should also not be a normal distribution. Please check your bins and extents (for example the dark blue and dark orange should be more prominent).

Please plot the histogram without preset bins so that you can see what it's supposed to look like and then follow the example in the original issue for visualizing the correct histogram. Please ask questions if you don't understand the code.

The new dataset isn't a normal distribution so the histogram should also not be a normal distribution. Please check your bins and extents (for example the dark blue and dark orange should be more prominent).

Please plot the histogram without preset bins so that you can see what it's supposed to look like and then follow the example in the original issue for visualizing the correct histogram. Please ask questions if you don't understand the code.

I took a look at the original issue but it says the new dataset is a bivariate normal distribution, should I still change the histogram? And I'm a bit confused about the extents, do you have any suggestions on what changes should I make?

I changed the histogram slightly so it depicts a bivariate normal distribuition, please tell me if I still need to fix anything

Did you look at the output?

https://output.circle-artifacts.com/output/job/28ecf74f-5057-416e-905b-29c996c5c4e1/artifacts/0/doc/build/html/gallery/color/colorbar_histogram.html#sphx-glr-gallery-color-colorbar-histogram-py

Yes and it looks slightly different from the output I got when I ran the program locally...does it have to do with the output screen scale?

I made some adjustments to the positions of the image and the histogram, it should look a bit better now, I also tried the changes locally and it works just fine!

I'd also figure out a distribution that had more features in the histogram. Also it is quite strange to have high values be blue and low values be red. (Yes I know the color map is upside down, but I'd just use the reversed version)

So sorry about the last 2 commits, I had some version conflicts in my branch and I was just trying to solve it. I made some changes suggested by@story645 and now it is showing the none loop version

I found something like

Z1=np.exp(-((X+1)*1.3)**2- ((Y+1)*1.3)**2)Z2=2.5*np.exp(-(X-1)**2- (Y-1)**2)Z= (Z1**0.25-Z2**0.5)

gave a much more interesting histogram.

I would also strongly suggestcmap = plt.get_cmap('RdYlBu_r') - blue for positive and red for negative is very unconventional.

The new histogram is in fact very interesting, should we keep it as the example?

The new histogram is in fact very interesting, should we keep it as the example?

I think it's cool, but what happens if you try like 21 bins? Basically, what's the minimum number of bins you need to have an interesting shape?

The new histogram is in fact very interesting, should we keep it as the example?

I think it's cool, but what happens if you try like 21 bins? Basically, what's the minimum number of bins you need to have an interesting shape?

I see...so should we increase the number of bins we are using or do you think it would be better to go back to the previous histogram?

The new histogram is in fact very interesting, should we keep it as the example?

I think it's cool, but what happens if you try like 21 bins? Basically, what's the minimum number of bins you need to have an interesting shape?

I'm not sure the context for this question. In general, you want as many bins as you can get without having too many singleton bins.

I'm not sure the context for this question. In general, you want as many bins as you can get without having too many singleton bins.

Just that for the visualization it would maybe be less busy if there are fewer bins but it's not a big deal/blocker.