Parallel Coordinates Plot and Sunburst Chart

Hey,

So I have the following dataset. and I want to create a parallel coordinates plot and a sunburst chart.

However, in the second column, I only want to select, one player for and then add the statistics. However, I can’t and it selects all the players so the chart is a big mess. How do I select an individual player?


may work for you or other Row Filter nodes.

Hey, I can’t seem to see the player names under the player column.

Try


If the number of players more than 60 you need to set the list size using

2 Likes

Hi there @adepani,

as I told you here I don’t think Parallel Coordinates Plot node can help you. On the other hand Sunburst Chart node might but prior to that seems you have to do some (heavy) preprocessing.

This is example table you need for Sunburst Chart:

Player Area Assist Number
Al Horford Within 4 ft Gordon Hayword 20
Al Horford Within 4 ft Marcus Smart 5
Al Horford Short 2 Terry Rozier 17
Al Horford Short 2 Marcus Smart 16
Al Horford Short 2 Jason Taytum 20
Al Horford Long 2 4
Al Horford Long 2 Marcus Smart 8
Al Horford Corner 3s Gordon Hayword 9
Al Horford Corner 3s Marcus Smart 20
Al Horford Short 3s Terry Rozier 25

Then you will get something like this:

But as @armingrudd asked here you should upload sample data (for example first 10 rows) if you want help manipulating your data to get format you need…

Do you need help with uploading sample data?

Br,
Ivan

2 Likes

Hi, here is a sample data.

2018_19_NBA_shot_logs_sample.xls (36.5 KB)

I want organise the left side & right side by counts, and colour the bars according to whether the shot was assisted.

For example for the picture below, The player on name is on the left, the zone is in the middle, and the assist, is connected to a plate name. The stuff in grey, is the # shots that were not assisted for each zone and they are connect to a grey column without a player name (on the right).

I need help manipulating the data so I can create sunburst charts to mimic the following.

Also, that is the type of preprocessing I need to organize my data like that. I can’t seem to get it like the table you have. I have attached the sample data in the post below.

Hi there @adepani,

data sample is too small for example workflow but here are preprocessing steps in order to have data in format from my reply in order to try show statistics you care about in Sunburst Chart :

Give it a try and if any problems/questions feel free to ask and share your workflow.

Br,
Ivan

1 Like

Also, a couple of questions about your list, by shot entries I selected event_type and then included shot and miss.

For color manager it won’t let me color based on assist.

And lastly, I am confused about what you mean by this: " add nominal value row filter and sunburst chart into a component?

Here is the screenshot of my workflow.

And here is another dataset with 10,000 rows.

Shot Logs Sample.xlsx (661.8 KB)

Hi there @adepani,

why include both shot and miss? If I got it right you are interested only in made baskets and those have label shot.

Regarding assist column in Color Manager dialog. If you have more than 60 unique values (which you probably do) you need Domain Calculator prior to it. Check here for more info.

Regarding Components it is just a suggestion. For more info check this video:

Hope this helps!

Br,
Ivan

Hi! So two quick questions.

I added the domain calculator before the color manager and connected it to the column resorter on the left and the color manger on the right. In it I unchecked the “restrcit number of possible values” and selected assist, player, shot_zone, and count*(shot_made). Is that right?

Also, How did you select a specific player for the sunburst chart? When I display mine its all the players?

And plus this is the type of output I am getting.Documents.xlsx (985.0 KB)

Hi @adepani,

Seems so. Only why adding shot_zone and count*(shot_made) columns?

I used Nominal Value Row Filter node.

Output seems fine :+1:

Br,
Ivan

Ok, I think I got it to work. The last thing I am confused about is, if I am displaying the total shots that weren’t assisted as well? Because in the parraell coordinates graph above, the great lines are showing the total # that weren’t assited in each zone.

And regarding this comment “Seems so. Only why adding shot_zone and count*(shot_made) columns?”

I am guessing in the domain calculator all I have to select is player and assist and leave out shot_zone and count*(shot_made).

1 Like

Hi @adepani,

glad to hear you are making some progress.

That is my understanding as well.

Yep. Cause both shot_zone and count*(shot_made) domain’s has less than 60 unique values and even if it has more it is not relevant to have all values in domain for those variables.

Once done don’t forget to come back to inform what grade we got :smiley:

Good luck!

Br,
Ivan

Thank you, so I am displaying the shots that weren’t assisted right for each zone for each player I select?

Hi @adepani,

seems so - if that is what you want.

Br,
Ivan

This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.