Sea Bottom Profile

[MikeArmstrong]

Hi

I have a query about Mathcad�s programming capabilities.

We have just surveyed a trench and have a set of x,y and z coordinates and wish to generate a 3D profile of the seabed.

Can this be done in Mathcad?

Will the 3D graph generated be of good quality?


Regards

Mike

[ValeryOchkov]

On 7/30/2009 2:39:00 AM, Armo wrote:
>Hi
>
>I have a query about Mathcad�s
>programming capabilities.
>
>We have just surveyed a trench
>and have a set of x,y and z
>coordinates and wish to
>generate a 3D profile of the
>seabed.
>
>Can this be done in Mathcad?
>
>Will the 3D graph generated be
>of good quality?
>
>
>Regards
>
>Mike

It is not seabed but see please
http://twt.mpei.ac.ru/ochkov/WSPHB/iapws95-97surface.PNG
http://twt.mpei.ac.ru/ochkov/WSPHB/iapws95-97counter.PNG
(Made in Mathcad)

Val
http://twt.mpei.ac.ru/ochkov/v_ochkov.htm

[MikeArmstrong]

Thanks for the reply Val.

That looks exactly what is required, how would I go about making a seabed profile say from x,y,z coordinates for 3m increments?

Regards

Mike


Regards

Mike

[ValeryOchkov]

On 7/30/2009 3:47:07 AM, Armo wrote:
>Thanks for the reply Val.
>
>That looks exactly what is
>required, how would I go about
>making a seabed profile say
>from x,y,z coordinates for 3m
>increments?
>
>Regards
>
>Mike
>
>
>Regards
>
>Mike

Do you have this x,y,z coordinates?
Val
http://twt.mpei.ac.ru/ochkov/v_ochkov.htm

[MikeArmstrong]

Yes I have the coordinates but there is alot of data.

We are looking at a 10m wide trench with left centre and right (sides of the trench) water depths for 3m increments.

Would this be enough data?


Regards

Mike

[StuartBruff]

On 7/30/2009 8:41:22 AM, Armo wrote:
>Yes I have the coordinates but
>there is alot of data.
>
>We are looking at a 10m wide
>trench with left centre and
>right (sides of the trench)
>water depths for 3m
>increments.
>
>Would this be enough data?


Does the attached help at all?

Stuart

[MikeArmstrong]

Stuart that looks like it's getting there, if you have a look at Vals first picture example that is exactly the sort of thing I am after.

The seabed readings will be every 2 or 3m, readings will be from one side of the trench to another for up to a thousand metres in length.

So I need a seabed profile from the data I receive from the surveyor.

Hope that�s not to confusing.


Regards

Mike

[StuartBruff]

On 7/30/2009 10:02:10 AM, Armo wrote:
Stuart that looks like it's getting there, if you have a look at Vals first picture example that is exactly the sort of thing I am after.

The seabed readings will be every 2 or 3m, readings will be from one side of the trench to another for up to a thousand metres in length.

So I need a seabed profile from the data I receive from the surveyor.

Hope that�s not to confusing.

Do you mean something like the format of the data at the bottom of the attached worksheet, Mike?

Stuart

[ptc-1368288]

That looks exactly what is required, how would I go about making a seabed profile say from x,y,z coordinates for 3m increments?
Do you have this x,y,z coordinates?
Yes I have the coordinates but there is alot of data.
So I need a seabed profile from the data I receive from the surveyor.
_______________________

Attach the data set in a Mathcad input table, "Save as" version 11
I will look at it for a "Waterfall" plot.

jmG
PS: I'm replying to your last visit "empty of the data set".

[MikeArmstrong]

Regards

Mike

[TomGutman]

Mayhaps you should post some of your real data. At this point it is very confusing. You talk about x, y, and z coordinates, normally indicating a surface plot. But in this sheet you have only two vectors, indicating a simple 2D plot. What, exactly, is your data and how do you want to display it?
__________________
� � � � Tom Gutman

[MikeArmstrong]

Tom

Sorry for the confusion, to be honest i am not sure what plot is required as this is the first time i have looked at something like this.

Firstly I recommended a( x y z) data but when I was inputting data for jMG i noticed that using zero as the centre line might me less confusing.

Regards

Mike

[RichardJ]

Perhaps what you need is two different plots? One would be x, y, z, in other words the entire trench. The other would be a cross section of the trench at a given y (so just x and z), equivalent to the data you posted?

Richard

[TomGutman]

Post the data. Then we can see what can be plotted from it. For generic information on 3D plots you can read through the help and the quicksheets. For detailed information on all kinds of graphics, get the e-book on creating amazing images in Mathcad.
__________________
� � � � Tom Gutman

[ptc-1368288]

... you are confused . You don't need (x, y, z) data, you need only the columns 'z' of the cross-sections as it goes along the trench. If you have 11 points measured for each cross-section the data array will be 11 x n where n will be the number of columns. For handling a long trench in viable sections, it would be wise to have several data sets so that things can be viewed on plot.
Never mind centered or not centered, the plot will be first on indices, then when it comes to the "metrics", it will be metr'd at will and as fine as desired for an export.

Resume:
A rectangular data table of 11 rows by 100 columns or all the columns you have in hand.

[MikeArmstrong]

Ok I sort of get it.



I have attached a copy of some of the data, is this enough as a starting point.



Regards

Mike

[StuartBruff]

On 7/31/2009 2:24:18 AM, Armo wrote:
== I have attached a copy of some of the data, is this enough as a starting point.

Mike,

You just need the depth data to do the plotting of the surface.

Stuart

If you want to display the actual width and length values, then you need to create the arrays as per my previous posting example.

[ptc-1368288]

On 7/31/2009 2:24:18 AM, Armo wrote:
>Ok I sort of get it.

I have
>attached a copy of some of the
>data, is this enough as a
>starting point.

Regards

Mike
______________________________

More columns, so that the work done can be extended. Plug your data in your original work sheet "Seabed" ... one project in one work sheet. Zap the demo part that was added in your "Seabed" in order to minimize confusion.

jmG

[MikeArmstrong]

The data has been inputted into the original sheet and attached

Regards

Mike

[TomGutman]

The description of the data here is a bit different. That can be plotted easily enough.

A question does arise as to why you want to plot it? What do you expect to do with the data and the plots?
__________________
� � � � Tom Gutman

[ptc-1368288]

3 more types of plot added below Marlett in XX(1503).
Waterfall plot inadequate ... No more data ?

jmG

[MikeArmstrong]

Those graphs are starting to look like what is required i will supply the total amount of data to see if it would be better to work with.

Regards

Mike

[ptc-1368288]

On 7/31/2009 9:45:50 AM, Armo wrote:
>Those graphs are starting to
>look like what is required i
>will supply the total amount
>of data to see if it would be
>better to work
>with.

Regards

Mike
_______________________

Sorry Mike, but you are not correct !
For 28 threads you haven't provided an image of what it should look like , and you still hide the data set. All in all and between collabs many hours vanish in the blue, hours detrimental to collabs who need real help.

Provide:
1. an image of the "look like"
2. a large enough data set.

jmG

[TomGutman]

Sorry, I can't make any sense out out your data.

You say the first column is the trench width. You have eleven rows, so the implication is that you have eleven trenches, each with a width. But the regularity of the numbers, and the presence of a zero, render that unlikely.

And what is a cross-sectional reading? Usually a cross-section would be a surface of some sort, with a shape and an area. What is being measured here?

BTW, Mathcad can read a fair number of data file formats. It is not usually necessary (nor desirable) to retype data into Mathcad. If your data is in any sort of reasonable format just post the data file and have Mathcad read the file.
__________________
� � � � Tom Gutman

[MikeArmstrong]

Tom the first column is the trench width, the zero indicates the centre line of where the pipe will be laid It doesn�t have to be zero it just needs to be in the centre of the data.

The cross section reading is just a term I used it�s basically the reading from one side of the trench to another, and this is repeated every 3 metres along the trench length.

With regards to the data file, the number of readings done by the surveyor with a hand help GPS is too much and not in a clear format. (Over the top)


Regards

Mike

[RichardJ]

On 7/31/2009 7:17:42 AM, Armo wrote:

Here's my variation.

>With regards to the data file,
>the number of readings done by
>the surveyor with a hand help
>GPS is too much and not in a
>clear format. (Over the top)

A hand held GPS to measure the sea bottom? I would have thought some sort of sonar would be rather easier.

Richard

[ptc-1368288]

Super cool your slice explorer, Richard !

Same but more attractive plots.

jmG

[TomGutman]

>>With regards to the data file, the number of readings done by the surveyor with a hand help GPS is too much and not in a clear format. (Over the top)<<

Why do you say too much? Mathcad can handle arrays with millions of elements quite easily. Files of a megabyte, or even sometimes a bit more, are not a problem for the collaboratory. And what is "over the top"?

You should start with as close to the raw data as possible. If your data is originally in terms of absolute coordinates (GPS suggests latitude/longitude) start with that as the data. If you have a nominal centerline for the trench, that should be a distinct piece of data. And don't round the data. Start with the full precision originally recorded.

Rounding just adds additional error to whatever errors you already have in the data. Any processing is likely to lose information. While you may want to do some form of processing prior to plotting, that processing should be explicit, and preferably in Mathcad.

As far as format goes, well, your data in Mathcad has not been all that great. Let's just look at the actual data and see whether Mathcad can handle it or not. With a bit of programming it can handle some remarkably complex structures (cf the Nist standard regression data sets).
__________________
� � � � Tom Gutman

[ptc-1368288]

On 7/30/2009 12:17:29 PM, Armo wrote:
>
>Regards
>
>Mike
_________________________

What's needed is the all columns of the 'z',
otherwise you aren't not going to have the bed profile in no kind of plot. You can't have a soup w/o ingredients. If you have only 11 points per cross-sections all along that trench, you might want to interpolate all columns but that will not solve the problem completely. The profile will best result from interpolating the all columns as a rectangular spline. From there, the suite is to mesh and that will give the "metrics" of the profile ... from which you will be able to display the all trench or selected sections.

Your project stops here w/o data set.

jmG

[ptc-1368288]

... eventually, with enough columns , you will have a similar plot, at least for a start and a resulting project as a master tool to use again and again. If you have 11 data points per scan (per measured cross-section), you can have about 1.4 million columns of collected cross-sections at 1 m spaced ... you can collect data for a trench about 1400 km long.

jmG