A tale of 2 tables

Author: Pierre

There are 2 types of users of table:

  1. One that quickly look at it to find its information and move on
  2. One that works extensively with the table

While most UI guides online recommend a lean design, without the clutter of a big table with numbers, I will explain how to design a simple html table for power users.

A. Who ? Why ?

If you want to design applications for:

  • accountants
  • meteorologist
  • or anybody needing to look at a lot of number

They will need to have all of them in front. That means that you need to think about how to display efficiently and nicely a table with a lot of data.

At that point, nice UI and neat presentation are not on the agenda, when you work 8 hours on tables and data, you need to be efficient and not to scroll endlessly to access what you want or worse, have to constantly filter all the data you are interested in.

This example of RefactoringUI is for me a bit too reductive, and can apply to quick view of your website or limited data interactions. Notice the increase in size of the table too:

refactoringUI table

This to me, is a perfect table for a website that doesn't require a lot of interactions with tables. Think of something like a reservation for a hotel or booking your flight tickets. But not for spending time on it.

B. FiveThirtyEight ft. Michael Jordan

Usually you will find big tables on specialised platform and enterprise software. However I found that https://fivethirtyeight.com/ has actually done the best job in terms of table design ! And for a a news outlet.

Here is an example from their website, which does a good job at big tables, look at cropped example in a Michael Jordan's article : Why Michael Jordan Was The Best

FiveThirtyEight Jordan

While here we are looking at a table of a news website, we can derive a good idea to build our own tables. This is enough for making something readable and with a lot of information.

Let's dive into our interpretation.

C. Do you want to build a table ?

I will expand from the FiveThirtyEight example and show my interpretation of a good table.

One of the few good features to display a lot of data is to merely... Have a smaller cells and letters, so I would recommend to reduce the font size of the table (use the rem unit to scale them depending on the size of your website font size):

table {
	/*Some other properties...*/
    font-size: 0.7rem;
}

In general I like the idea of having the head of the table as uppercase words and maybe slightly bigger to make them stick out better, so nothing too complicated, we can just style the head of the table with this:

th {
    /*Some other properties...*/
    font-size: .8rem;
    text-transform: uppercase;
}

To top it, here, nothing too fancy, just a hover with a slight alpha layer on the table rows. We want something to be applied only to the body of the table and have neat table rows. And voila:

tbody>tr::hover {
    background-color: rgba(155,155,155,.2);
}

And here is our table:

Date Open High Low Close Adj Close Volume
2020-03-19 19900 20393 18834 19853 19853 239922
2020-03-20 19583 20882 18825 19875 19875 285733
2020-03-22 18504 18507 18086 18238 18238 204991
2020-03-23 18230 19804 18088 18745 18745 119535345
2020-03-24 18987 20943 18925 20511 20511 106863898
2020-03-25 20446 21892 20279 21100 21100 130872632
2020-03-26 20998 22545 20510 22492 22492 123364604
2020-03-27 22123 22321 21316 21410 21410 98666789
2020-03-29 20926 21244 20778 21213 21213 711107
2020-03-30 21350 22250 21153 22196 22196 82143599
2020-03-31 22204 22386 21470 21502 21502 66006634
2020-04-01 21397 21423 20648 20932 20932 82218631
2020-04-02 21030 21374 20500 21160 21160 97616148
2020-04-03 21044 21331 20740 20928 20928 63241057
2020-04-06 21647 22663 21620 22555 22555 96739993
2020-04-07 22496 23477 22300 22488 22488 274505
2020-04-08 22498 23426 22337 23417 23417 98632961
2020-04-09 23267 23889 23020 23676 23676 110699818
2020-04-12 23839 23911 23108 23331 23331 816363
2020-04-13 23295 23609 22968 23368 23368 66333180
2020-04-14 23621 23967 23534 23843 23843 75910145
2020-04-15 23757 23809 23103 23211 23211 81277443
2020-04-16 23265 24327 23079 24157 24157 94173258
2020-04-17 24239 24305 23701 24151 24151 105186538
2020-04-19 23944 24028 23917 23970 23970 284615
2020-04-20 24143 24167 23441 23530 23530 87221700
2020-04-21 23273 23433 22814 22999 22999 111731192
2020-04-23 23315 23772 23233 23282 23282 78882561
2020-04-24 23246 23460 23166 23416 23416 9038285

You can find an example in React and css at this repo: blog-posts.

You can browse the table, identify easily the rows and have plenty of data to display to your users. I think it is a great addition to the components a website could have, and especially if you design something for data savvy people.

I will expand soon on how to add interactivity with the table like pagination and filtering with react-table.

Both tables should exist, though you should strive for understanding what is beneficial to your user. Are they quickly looking for information or looking for deeper knowledge or investigating a situation ?

Now some news about the site

During the second part of 2019 I managed to build this website, piece by piece, to have a look that would suit what I want and some functionalities that I sought for.

One of the main goal is to cover basic statistics and analytics, common programming tasks encountered in life. But above all it is a place where I want to journal issues that I found interesting or small applications that I think could be beneficial - at least - for me and perhaps for others.

Over the Christmas holidays I started to create a small tool to cover basic exploration for data analysis: which techniques are the most useful to uncover hidden gems in your data.

While the tool covers broad scenarii, it doesn't do everything needed for an exploratory analysis. It is rather a good bookmark when you are not sure how to approach a statistical problem.

You can find it at https://www.chevallier.dev/statistics/ and it works by selecting the type of variable you would like to analyse (and it has some of the features of this "big table" !).

I'll make a comprehensive guide in the weeks to come and shall enrich piece by piece.

Copyright © 2019. Pierre Chevallier
Check my GitHub repo git logo