How to to connect between each value of point A an point B

I have two points with different coordinates point A=x,y for 1400 numbers and point B= x,y for 14000 numbers.

First: i want to connect between each value of point A an point B, Second: I want to get a new variable which contain in the first row the first value of x for point B and in the second row the first value of x for pA and apply this for all values from two poins: to make th extrapolation form this values to predicte any values from this line

please look at the attach pictuer Thank you for help

[Jan]

The question is not clear yet. Di you have one point or 1400 points? If you want to connect all of the 1'400 with all of the 14'000 points, you get 19.6 * 10^6 lines. This will exhaust the display and look like a fat stain only. I do not understand your second point. The relation between pB, pA and the prediction point is not clear. If you have two sets of points, there is no unique definition for a "prediction".

[Mansour Attaallah]

thank you for reply. yes, it's 1'400 with all of the 14'000 points. The relation between pB, pA it comes later when i do extrapolation but first i have connect these point to get new x and y axis whic contain the coordinates of the all lines

[Jan]

The question is still not clear. How do you want to use 19.6 million "connections" to extrapolate what or to "get new x and y axes"? Are you looking for a PCA (search this in the net)? Do you mean the convex hull of the points? I cannot imagine why creating a huge list of connections is useful. And in addition tow sets of points do not allow to "extrapolate" or "predict" anything - a substantial part of the problem has not been mentioned yet. An extrapolation requires to choose a model to describe the existing data, and a fixed component, at which the model is evaluated. Maybe you mean a line through the centers of the sets of points and you want to find the point, where this line crosses the X axis. But this is a bold guess only.

[Mansour Attaallah]

Thank you very much. you are right its too much. a friend of me fix this and made a new calculation. a regression to identify x axis from y axis to detect the cross section point with horizontal line. And present one graph to make sure that his calculation is right