The fact that the discrete pmf's are rendered using an underlying line chart structure leads to some aesthetic oddities...

This can be fixed by using a bar chart structure, where point masses are depicted using vertical line elements.

I'm not sure that having a third set adds any value to the visualization. It might be better to keep things simple and stick with 2 sets.

...as suggested in the Slack

Clicking a slider location moves the slider but has no effect on the displayed pdf/pmf or cdf. The slider must be slided for the proper effect to occur.

Should use parameter lambda for exponential not 1/lambda so that parameter is the mean.

It might be helpful to add a picture of the current state of the bag with the remaining marbles.

I think the portion of approximating pi may be a bit misleading. Perhaps a sentence or two could be added, specifying that pi is not a rational number and that you are using probabilistic methods to approximate it with fractions.

How could we talk more about regression with categorical data for either the independent or dependent variable? Is there a way we could bring up Linear Discriminant Analysis, Probit Regression, and Logit Regression?

How can I translate the seeing theory webpage (like existing ones in es and cn)?

User Story: As a learner of statistics, I would like to read and seeing theory in the language I speak best
Contributor Story: As a speaker of both English and an other language, I would like to translate seeing theory

problem for opening the follow link:

https://cs.brown.edu/courses/cs022/static/files/documents/sets.pdf

on

Chapter 2
Compound Probability
set theory

"forgotten a square in the variance definition part C on page 12 of the textbook"

Something broke the graphics in the last build of the docs/ dir

The old PDF has graphics: https://github.com/seeingtheory/Seeing-Theory/blob/625f7ae82aa8cc2e5fa3c3973be8d96188fbcab9/doc/seeing-theory.pdf
But latest PDF is missing graphics
https://github.com/seeingtheory/Seeing-Theory/blob/master/doc/seeing-theory.pdf
also http://students.brown.edu/seeing-theory/doc/seeing-theory.pdf

Resizing bar charts causes vertical scroll, especially on mobile.

Page 12
Proposition 0.0.1
(c)
Proposition do not match the equation obtained in the proof @page 13.
Proposition is Missing a square sign in the end.

consider readjusting the text to read and flow as follows:

“...But as the number of flips increases, the long-run frequency of heads is bound to get closer and closer to 50%. For an unfair or weighted coin, the two outcomes are not equally likely. You can change the weight or distribution of the coin by dragging the true probability bars (on the right in blue) up or down.” [Move this text here]

[Buttons]

"If we assign numbers to the outcomes — say, 1 for heads, 0 for tails — then we have created the mathematical object known as a random variable.” [Leave by itself]

Won't fix

Currently there are 5 discrete distributions and 8 continuous. Should we limit the number of distributions a user sees while also giving an option to see more?

The label n should be r in the table in combinatorics viz.

In most of the visualizations that use tooltips, the tooltips are not placed where they should be.

The definition of event B = {second flip heads}, however the set of outcomes that follows, {HT, TT}, represents {second flip tails}.

See screenshot where slider background (white line) is missing:

Looks fine in Chrome.

There is just a small translation mistake in the Chinese version in chapter 2 section conditional probability. It says 在今天多云 的情况下，我们会估计“明天下雨”的概率小于“今天下雨”。whereas I suppose it should be在今天多云的情况下，我们会估计“明天下雨”的概率要高于一般情况下“明天下雨”的概率。

From a UI perspective, it might be better to change the sidebar button (the square with the three horizontal lines) to a dropdown menu, because

1. A dropdown menu can be titled with text, so the user has an idea of what the menu will contain before opening it.
2. Scrolling is disabled while the sidebar menu is open. The user must click the icon in order to close the menu and re-enable scrolling.
3. In my own experience, I generally find expanding sidebars annoying unless I'm on a mobile interface.

In your Chapter 3, point 2, you have F(x) = P[X < x].
The usual definition for the cumulative distribution function is F(x) = P[X <= x].
In the continuous case, we have P[X < x] = P[X <= x]. But in the discrete case these two probabilities are not equal. For example, for the case where X has a Bernoulli(1/2) distribution, P[X < 0] = 0 and P[X <= 0] = 1/2. It would be nice if you would use the standard definition F(x) = P[X <= x].

On the "Variance" section in "Chapter 1: Basic Probability" the number three card has a 4 instead o 3 in the bottom right corner.

Changing the bar graph labels caused the coin images to overlap with the axis text. The margins/padding on the coins needs to be adjusted as a result.

Link is broken on homepage for feedback on the book.

It could be a good idea to add a link to the header that takes the user directly to the landing page. right now the user needs to open the menu on the right-hand side in order to navigate home. The home page link could occupy the space where the unit title is placed in the top left. Clicking on the unit title currently just scrolls the user to the top of the page. Alternatively we can keep the unit title and add a "Seeing Theory" link that takes the user home.

Ability to reset each example to redo it without reloading the page.

It seems to be broken for all languages:

The link for English is: https://cs.brown.edu/courses/cs022/static/files/documents/sets.pdf
It returns 404 error code.

1. Limit current distributions allowed for confidence intervals and add discrete options

• Discrete (Binomial, Poisson)
• Continuous (Normal, Uniform, Exponential)

2. Add option to create credibility interval or confidence interval

• Add conjugate prior
• Change estimation panel from averaging to showing posterior distribution

The Set Theory visualization might be improved by labeling the circles (A, B, C) directly as opposed to providing a legend on the side.

"In the section on Variance in ch. 1, the green square to the right indicates the sample variance. What is confusing about this example is that the variance is calculated empirically, but it uses the analytically derived mean, rather than the sample mean. It seems like the example would be more consistent if both the mean and variance were estimated based on the sample. I realize this may be hard to visualize, but I thought I would let you know."