New to Dataiku DSS? Try out our NEW Quick Start Programs today and get onboarded on the product in just one hour!

# univariate analysis of several variables in one graph

Solved!
Level 3
###### univariate analysis of several variables in one graph

Hi,

I have a table like below. DSS can show the variarite analysis (like boxplot)  in the 'statistics', but only seperately. Can I create a graph, which include three boxplots together, so that one can clearly see the comparision among those variaibles.

Thanks a lot!

 work hours 2017 work hours 2018 works hours 2019 42 40 42 41 45 43 45 40 41 43 42 40 42 41 41

1 Solution
Neuron

several ideas come to mind

1. Restructure the data in long form, with a column for years and a column for hours.  In the R world this is some times called tidy data.

2. Use dashboards and post the three charts you already know how to make to a single dashboard.

--Tom
3 Replies
Neuron

several ideas come to mind

1. Restructure the data in long form, with a column for years and a column for hours.  In the R world this is some times called tidy data.

2. Use dashboards and post the three charts you already know how to make to a single dashboard.

--Tom
Level 3
Author

Thanks @tgb417 the first option is good! 🙂

Neuron

Excellent!

Long form or so-called "tidy data" can be super helpful for graphing.

Knowing how to use the Visual Prepare recipe to "fold" multiple columns data can be very helpful in creating tidy data set in DSS.  Here is a similar example I've been working with recently related to a number of customer name columns.  Where I want the column name in one column and the names in another.  (You can think about this a "unpivot" function.)

In your case in the column list, you would put each one of your columns in your table.

• "work hours 2017"
• "work hours 2018"
• "work hours 2019"

--Tom