Achievable things in AB (using a tiny bit of imagination)

[fxshrat]

Building upon previous post's example of using matrices I have managed to plot the efficient frontier using exploration.
This is just to show you that you don't necessarily need to use Gfx...() functions for plotting such result data.
Of course using Gfx offers more flexibility (IMO). But for quick plots XYChart...() charting functionality is a useful addition within AB.

(other table output examples amibrokerforum.proboards.com/thread/14/addrow-function-vrs-5-91)

Table:
i.imgur.com/JbXBvLJ.png

[fxshrat]

As for VAP (volume at price)...

In case you are not aware of it yet but with help of new function PriceVolDistribution of AmiBroker 6.20 development cycle you can calculate Volume profile on lower interval and not only output it on that same lower interval (i.e. tick or 1-minute) but you can store it there (persistent variable feature of static vars) and then in addition you can output the data in any higher interval (i.e. daily time frame) too.

See picture example. Data has been calculated on 1-minute time frame and for the output of the data daily time frame has been chosen. So again what you see there as example is 1-minute VAP on daily bars.

[fxshrat]

With further reference to this thread amibrokerforum.proboards.com/thread/159/volume-price-buying-selling-volumes?page=3 here are just a few other examples of order flow analysis similar to GomOrderFlowPro etc. using AmiBroker (once again silencing inexperienced "experts claiming it would not be doable in AB at all. Oops...)

Examples show footprint charts with dominant (diagonal) side modes (or volume imbalance mode).

[fxshrat]

Communication between AmiBroker and 3rd party programs:

For example sending analysis result list to Excel directly


Sending charts to Excel directly


Sending data from Excel to AmiBroker to output


A basic example of AmiBroker communication to R

[fxshrat]

Here some combination of Scatter, Linear Regression an Boxplot (plots either external or internal data). So such plots are pretty much independent.
Since one can store data to multi-dim array one can easily scroll through the number of available columns via Param().

It is similar to R-scatter plots such as to be found here www.statmethods.net/graphs/scatterplot.html
Difference: In AB it is easier with better+faster data access there (IMO).



And corresponding 'Residuals' plot (as well as statistics summary such as t-value and p-value)