Issue with creating cross-row relationship with plugin, and

For reference, I have tried the 7.5+ version of the plugin with Kibana 7.5.2 (and 7.5.0,7.8.0), along with the 7.0-7.4 version of the plugin with Kibana 7.0.0 (and 7.4.0).

I have an index pattern formed by a simple CSV as follows:

```csv
A,B,C
1,1,1
1,1,2
1,2,3
1,2,4
2,3,5
2,3,6
2,4,7
```

My goal is to form a graph with the edges connecting the nodes representing the 1 from column A to the 1 in column B and 2 column B, and the 2 in column A to the 3 and 4 in column B, and the 1 in column B connected to the 1 and 2 in Column C, the 2 in Column B to the 3 and 4 in column C, the 3 in column B to the 5 and 6 in column C, and the 4 in column B to the 7 in column C.

Based on the screenshots and instructions that have been provided in the website and the User Guide, it seems that would be possible, but I am having trouble being able to actually materialize the visualization.

(Let A be CUID, B be DUID and C be RUID in the following screenshots).
Let us consider only trying to achieve the relationships between columns A and B first.

So, the only aggregation/sub-aggregation options I get are "Terms" for this dataset.

Here is a look at what I get using a Node-Node relationship and Count aggregation
The big blue circle is labeled as "CUID:1 Count:4", the small one is "CUID: 2 Count: 3", the orange ones say "DUID:3" and "DUID:4".

Here is a look at using a Node-Relation model:
The big blue circle is "CUID 1: Count 4" and the small one is "CUID 2: Count 3"

and as you can see adjusting the parameters in Relation to use the aggregation of minimum for C seems to have no impact...
The big blue circle still says "CUID 1: Count 4" and the small one is "CUID 2: Count 3"

Here's a look at using the Node-Node with an Average aggregation:
The big blue circle says "CUID:2 Average DUID: 3.333333", little circle says "CUID:1 Average DUID: 1.5", orange ones say "DUID:3" and "DUID:4".

Any guidance will be kindly welcomed.

Start by reproducing the CSV example in the Kibana plugin using the Node-Node and Node-Relation models, first with Terms and then with Count, minimum, and Average aggregations. Compare the resulting graph with the requested A-to-B and B-to-C edges and verify whether changing the Relation aggregation changes the visualization. Done means the configured relationships and aggregation values are reflected in the graph, or the limitation is documented.

Written by the indexing model from the issue text.