I'm using NodeXL to plot a lot of points which are actual coordinates for cities.
The thing is that the way it's plotted now is that North America is on the Bottom Right but it should be on the top left like on a normal map.
It's like this for all of the points so pretty much I need to rotate the whole graph so that what's on the bottom right will be top left after transformation, and what's on the top right will be on the bottom left.
I have two columns with X and Y points as follows (for example):
X Y
6,238.2 9,896.0
6,141.9 9,896.0
I'm not sure the formula or Math behind this kind of rotation.
The graph is only positive so from (0,0) upward and outward to the right, there are no negative values on the x or y axis.
Could anyone help me out?
For the sake of an answer:
To quote #Tim Williams: Instead of plotting x and y plot (width-x) and (height-y)
Example:
Related
I have a straight trendline on a semi-log chart (ie y axis in log scale, and x axis (ie time) in regular scale).
I think therefore the equation of that straight line is y = a * b^x
I want to move the line up or down though according to trigonometry but don't know if I can as the opposite side of the triangle is in log scale. I can do this fine on regular scale charts, but I dont know where to begin for semilog charts.
Anyone know how I do this or if there's an equation I can use to calculate how much i move the line up or down?
The following code
a= np.random.randn(100)
b=np.random.randn(100)
sns.jointplot(a,b)
produces
What is the histogram on the top and right hand side?
And how do I read this graph?
The histogram on the top shows the distribution of the variable at the x-axis and the histogram to the right shows the distribution of the variable at the y-axis.
And the scatter plot is the usual x=a Vs y=b. Hope it helps.
Dataser a's x axis also becomes x axis of the joint plot, database b's x axis becomes y axis of the joint plot. Graph on top and rhs, are same as the individual histogram. The graph is continuous, and changes on bin change, unlike histogram where it may be discontinuous
I've made a chart with Excel 2010's "Combo" option for chart type so I can plot two data series for the same depth points. The image below shows the default, which is very close to what I want - except I would like to have the axes flipped so that the current X axis, which is depth, is displayed as the Y axis and both primary and secondary current Y axes plot as X axes. In other words, I'd like to rotate the chart area by 90 degrees clockwise. The "Switch Row/Column" doesn't do what I want (or expect) and I'm running out of both ideas and patience. Is there an easy fix? Or a hard fix?
Here's the plot as-is:
And here's a dummy plot of the end goal made by rotating the image around in Paint, in case the picture makes it clearer:
Finally, as was pointed out in the comments, the whole thing looks goofy and might be better plotted as a bar graph with two bars. I tried this as well and came away almost all set - but the gray bars plot from left to right and the blue bars plot from right to left. Seems like it should be as simple as changing the "Plot Series On" option to Primary Axis for both, but this destroys the graph.
I looked around and I think this link has instructions for what you're looking for: https://superuser.com/questions/188064/excel-chart-with-two-x-axes-horizontal-possible
I'm experimenting with a vector based graphics style with objects represented as series of line segments with a given width(it would probably be easier to think of these as rectangles). The problem is that these segments are connected at the center and leave a gap (shown below). I've determined that the most efficient way to cover this gap is simply to cover it with a triangle, and since I'm working in OpenGL, all I need are the points of the two points that don't overlap with the other rectangle, the third point being the center point where the two line segments(rectangles) are connected. How can I determine which points I need to use for the triangle, given that I have all of the points from both rectangles?
EDIT: I will also accept alternative solutions, as long as they cover up that gap.
EDIT 2: Nevermind, I solved it. I'll post code once I have better Internet connection.
Maybe I'm misunderstanding the question... but if you zoom in on the top corner of your red pentagon, you get something like this, am I right?
where A and B are nodes on the rectangle for edge1 and C and D are nodes on the rectangle for edge2. You say you already know these coordinates. And from what you say, the edges meet at the centre, which is halfway between A and B, and also halfway between C and D. So call this point X, and you can calculate its coordinates easily I guess.
So all you need to do is draw the missing triangle AXC, right? So one way would be to determine that A and C are on the "outside" of the polygon (and therefore need filling) and B and D are on the "inside" and therefore don't. But it's probably easier to just draw both, as it doesn't hurt. So if you fill AXC and BXD, you'd get this:
The solution I found assumes that there are 3 basic cases:
First, the three unique center points for the two rectangle proceed upward (positive y direction) so the gap is either on the left or right of the connection. In my code, I had the corner points of the rectangle organized by their orientation to the left or right of the center point, so if the bottom rectangle's left point is below the top rectangle's left point, then the gap is between the left points of the two rectangles, otherwise the gap is between the right points.
Second, the three unique center points have a maximum at the center most of the center points, so the gap is on the top. The gap is then between the two points with the maximum y values.
Third, the three unique center points have a minimum at the center most of the center points, so the gap is on the bottom. The gap is then between the two points with the minimum y values.
[I'll post pictures of the example cases if it is requested]
How do I mark knot points at specific points on the curve not all. For example, I plot the following curve -
(x1,y1)
(x2,y2)
(x3,y3)
(x4,y4)
(x5,y5)
I want to mark knot points on x2 and x4. Is there some way to do this?
Make two data series, one with all points, another with the points you want marked. Format the first (all points) to display the line without markers, and the second (selected points) to display markers without the line.