swift-algorithm-club | Algorithms and data structures in Swift | Learning library

In terms of the basics of how to draw a line, you:

1. Create UIBezierPath;
2. Move to the starting point with move(to:);
3. Adding line to the end point with addLine(to:);

You can then render that path in your UI by either:

• Create CAShapeLayer, specifying its strokeWidth, strokeColor, and fillColor; set its path, and then add that shape layer as a sublayer of the view’s layer; or
• Create UIView subclass and in its draw(_:) method you can call setStroke of the desired UIColor, set the lineWidth of the UIBezierPath, and then stroke() the UIBezierPath.

Generally, I’d use the CAShapeLayer approach, where I basically configure the shape layer, but let the OS render that shape layer for me.

That having been said, I’d probably take this a step further, and wrap the line drawing in its own UIView subclass. The thought process is that not only are high-level views generally composed of UIView objects, but it also opens the door for all sorts of advanced UX (e.g. you might want to detect a tap on a node and do something).

Anyway, I’d wrap that “connector line” drawing code in a UIView subclass, like so:

Try this out:

My solution is a little bit more wordy :) where I replaced printTree with my own two methods

Map your generic type T to String and then apply joined() function.

joined is only for elements of type String and here compiler do not know the type of T.

I confirm that your bearing function works correctly. However, the regression value you get for latitude is 51.45437262420372 - which is where the regression line intersects with longitude = 0.0. I think you need to use both the intersection point and slope of the regression line to calculate fitted lat,lon values (within your map region), and then find the bearing using those values.

LATER ADDITION: An important issue here is that longitude coordinates are not linear in relation to distance - the distance covered by a degree of longitude varies with latitude. I think the simplest solution is just to convert your array of lat,lon values to lat,normalisedLon values where normalisedLon = lon * cosine(lat).

When you do this the angle of your path from your first coordinate to your last coordinate = 3.055 degrees (just east of north which appears to be correct).

Once you've done this then the slope of your regression line will in fact represent the direction you are seeking i.e. heading = atan(slope).

However, your regression code just doesn't seem to work for me. It should give a result very close to 3 degrees also, but it is way off.