This page on the Mathematica StackExchange contains some 2016 code for the PlotExplorer which István Zachar began writing for Mathematica 9. The old code had problems, but this code seems to actually work.

http://mathematica.stackexchange.com/questions/94716/how-to-use-a-ready-function-in-your-code-for-panning-and-zooming-on-2d-plots/94759#94759

It really provides a variety of functions. To try it, cut and paste it and add this:

PlotExplorer@ ListLinePlot[Table[Accumulate[RandomReal[{-1, 1}, 250]], {3}], Filling -> Axis]

Some interesting demonstrations is here along with a 2015 version of the PlotExplorer. I was unable to make this version work in MMA 10.0.0, late 2009 iMac.

http://mathematica.stackexchange.com/questions/7142/how-to-manipulate-2d-plots

In the link you provide in your post above, István Zachar says

When it comes to visual analysis, large datasets or data with intricate internal details often makes plotting in 2 D useless, as the outcome is either just a fraction of the full dataset, or no details can be observed in the mess of datapoints.How can one make the process of changing the plot range and/or zooming, panning, etc.less tedious than doing it programmatically and iteratively, from time to time?I often meet with this issue, and developed various methods to deal with it (like this).Though I have now a working solution that I would like to share (see below), I also am highly interested in what kind of methods and tricks others invented to visualize and manipulate complex 2 D data with ease.

I fully agree with the remark, and have spent a lot of time struggling with the issue for years..

It seems that in Mathematica V9 WRI included an experimental function (PlotExplorer[]) that disappeared in versions 10 and 11.

Any comment will be welcome if anyone has made any successful use of the István Zachar' sPlotExplorercode published in the Mathematica StackExchange post linked by Gary,

@Gary

The first one resets it's max value each time you drop the slider.

Maybe this one is more obvious, where the slider width changes with values of x:

Clear["Global`*"]; 
x = 1; 
{Dynamic[Slider[Dynamic[x], {0, 1}, ContinuousAction -> False, ImageSize -> x*1000]], Dynamic[x]}

Versus this one where only the initial value is used and the slider isn't changed after evaluation (although the value of x is)

Clear["Global`*"]; 
x = 1; 
{Slider[Dynamic[x], {0, 1}, ContinuousAction -> False, ImageSize -> x*1000], Dynamic[x]}

Quotes have are preceded by an ">" sign.

Code snippets are preceded by a tab indentation.

Click on the icons at the top of the box where you write the messages. The left most one is for code, the right most one in the second column is for quotes.

@Patrik, That is what I was looking for. Thanks.

But what about the explanation? Let me pose a few questions:

1. AutoAction is suppose to enable the controls to take action whenever the pointer is over them, even if they are not clicked. I suppose in this script, that must mean whenever the pointer is over the locator pane, which has the dimensions of PlotRange. Is that right? What is the role of AutoAction in this script? What action are the controls taking when the pointer is over the locator pane (as opposed to actually being clicked)?

2. The documentation for AutoAction says that it is an option for controls such as Slider, Locator, and Button, yet you have written it as an option for LocatorPane[]. Is that because LocatorPane[] is also a control like Locator?

3. Since you have Dynamic[pt2] inside LocatorPane[], what is the need for Dynamic[LocatorPane[]]? Or perhaps I should ask it the other way around.

4. And how is it that one can reset with right click?

@Gianlucca, We now have two such attempts. Thank you for making a try. Perhaps the traffic will inspire more experienced event handlers to jump in.

I noticed that you included "Graphics" as the argument to MousePosition[]. As I read the documentation on ?Graphics, this should yield the position of the cursor when it lies within the boundaries of the Graphics[], which perhaps means any position within the PlotRange rather than within the graphics objects (i.e. the rectangle or disk) that are the arguments of Graphics[]. I have not seen this explained. At least, when I include "Graphics", the behavior changes to provide what I was attempting. That is, the disk moves from point to point instead of disappearing. So that is a big improvement.