gtl | A collection | Build Tool library

You have done everything right except for that extra loop for the first iterator. Following is the complete working solution.

This is just a basic example how you could build the tree from a JSON response. You may have to adjust criteria to map child to parent (it's a depth-first traversal opposed to my originally suggested breadth-first).

Note that if you know if and how the JSON response is ordered you can safe significant time by eliminating the searching of the root node and it's related nodes.

MainWindow.xaml.cs

One way to duplicate the rows in a dataframe is to concatenate the dataframe with itself:

Controller action returns IActionResult, but you are comparing it to Title object. You should instead check if it's OkObjectResult and validate it's value:

Make the Tick handler async and await the long-running call:

The union of two polygons isn't strictly always a single polygon (what if they were disjunct to begin with).

If you know this to be the case, I'd suggest just getting the first polygon of the set (perhaps asserting that the set has size 1).

I think any planar geometry with a zero area is by definition self-intersecting.

Geometry algorithms often deal badly with those and the preconditions are likely documented. At this point I can't find such documentation with Boost Polygon/GTL.

From what I gather reading the source code, some overloads/variations of contains accept a boolean argument consider_touch that might help you out. Beware of unspecified behaviour though. (E.g. the polygon/polygon implementation counts intersections, and counting intersections with overlapping edges is going to be... questionable).

I'd use a segment/line for non-planar geometries, so that you don't startle any of the (undocumented?) geometry invariants or algorithm preconditions

Most Instruments have remote and local commands. Visa also has its own set local command, which looks like viGpibControlREN( handle, VI_GPIB_REN_ADDRESS_GTL ) in c. You'd have to read the manuals for the equipment to be sure, but their individual visa commands are usually something like "SYST:LOC"

It seems you are working on a regression problem (i.e. predicting continuous values). There are, at least, two things you need to consider:

1. As @Mitiku has mentioned in the comments section there are some NA (i.e. missing) values in the data. This is one of the reasons that makes the loss to become nan. Either drop the rows which have NA values, or alternatively replace NA values with a specific value such as 0. See this answer for more info about dealing with missing data.

2. Using accuracy as the metric for a regression problem does not make sense as it is only valid for a classification task. Instead use a regression metric such as mse (i.e. mean squared error) or mae (i.e. mean absolute error).

Please apply the two points above in your code, and then report back how the training goes, and I'll update this answer as needed.