Multi-level treemap in Plotly

Question

I am trying to build a treemap in my dash application. My dataset has a hierarchy that does not have a definite level. The original dataset is as below:

The levels of defined after splitting the code by "." from the Code column. After separating into levels, the dataset becomes as displayed in the picture below:

Now if I try to create a map with this

fig = px.treemap(df, path=['level_1','level_2','level_3','level_4','level_5','level_6',level_7'], branchvalues="total")
fig.show()

I get an error describing that non-leaves rows are not permitted

But, when I put a placeholder value instead of NaN's, I am able to plot the treemap.

At first the treemap seems ok. But, if we drill down the hierarchy, we see that structure in the treemap is not entirely correct. There should be totals lowest level columns, but what we see is the nested placeholders values 'tmp', as shown in the picture below.

Am I doing something wrong in building the dataframe structure?

Answer 1

I had the same issue! It seems like plotly treemap requires rectangular dataframe. So, when you have None value in various rows/columns; it makes it non-rectangular. Replacing None with df.fillna("tmp", inplace=True) solves that issue.

Reference: plotly.express.treemap