|
Their is no simple way for you to accomplish this. If you wish to do this you will have to create a loop that will search 2 layers into every folder searching for the specified file. It is much easier to accomplish if you have the exact filename/file path.
Regards,
Thomas Stockwell
Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning.
Visit my Blog
|
|
|
|
|
That "dot" notation does not work how your example implies it does. The dots just mean one of two things. A path starting with a single dot, ".\" means "current directory". A path starting with two dots, "..\" means the parent directory to the "current directory".
In your case, the path starts at C:\work, then goes up one directory, to C:\, then up again, to C:\, then specifies the file, File.txt. The destination is also in the root, C:\Fil007.txt. So, what you essentially did, was rename the file.
Using "dotted notation" like this is not recommended because of the assumptions made by "current directory".
|
|
|
|
|
I have a chart control. But what I need is a powerful chart control.
It should show about 14.5 million points without performance problem.
Are there any chart control, that can show 14.5 million points ?
I try to use normal chart control, it takes a long time to refresh.
I try to use directx to draw lines from 14.5 million points, it takes
also long times.
Does anybody know how to handle 14.5 million points ?
|
|
|
|
|
I think you have a design problem, placing that many points on a chart is well pointless. Surely you can summarise the information to a smaller dataset without compromising the accuracy of the display.
Never underestimate the power of human stupidity
RAH
|
|
|
|
|
I disagree. I've run into a similar problem doing geostatistics. Try this scenario:
You have 1000 well logs with readings every 1/4 foot and an average interval of interest of 1000 ft (about 4000 points per well and about 4,000,000 points all together). Now you want to calculate the vertical variogram, so you define lag categories every 1/4 ft, giving lags of 0.25, 0.5, 0.75, 1.0, etc. up to the maximum, say 1000 ft, or about 4000 categories.
To calculate the variogram, for each lag category, you take all possible combinations of points in each well that are separated by the given category distance range, sometimes the categories overlap, so there may easily be more points than you might expect, calculate the square of the difference, sum them all, then divide by the number of points in each lag category. It's a standard calculation. That, of course gives you an average variance vs. lag distance, the definition of a variogram, which is very smoothed.
Now you want to see how much smoothing has been done and if there are some outliers that may be screwing up the calculations or represent bad data. The standard means of doing that is by plotting the variance of the individual points in a variance cloud. The total number of points is about 1000 * (4000 + 3999 + 3998 + ...) = 1000 * 3000 / 2 = 2,000,000 points. You graph the variances of each combination of points vs. lag distance to generate the variance cloud, so you end up having around 2,000,000 points on the graph.
Of course, that's all well and good, except when I find a strange outlier point, I want to know where it came from, so I need to be able to identify which points came from which wells at what combination of depth, so I can go back to the raw data and see if it makes sense.
In this example it was only 2 million points, but I can easily see where you can get many, many millions if you go to overlapping lag ranges, where point combinations may be in more than one lag range, and/or three dimensions rather than two.
To do that I use my own graphics routines and some heuristics to try to figure out which areas of the graph will be totally covered with points and only draw them once. In the case in question, I have no idea if that's possible. Another potential trick is to do the calculations only once (fast) and save them in memory (fast) so you can generate the graph without recalculating anything. That of course involves a trade-off between memory and speed, which may or may not be an option. If you start swapping memory to disk, it may be even slower.
Unfortunately, I haven't found a great way to speed the point selection from the cloud up. My approach is to select an area on the cloud plot, then search through all the points to find out which ones fall within that area. If it has to hit the database every time, it's just slow, but it beats the heck out of doing it by hand!
CQ de W5ALT
Walt Fair, Jr., P. E.
Comport Computing
Specializing in Technical Engineering Software
|
|
|
|
|
Walt Fair, Jr. wrote: My approach is to select an area on the cloud plot,
I'm not debating the need for multi million data points I have an issue with attempting to display them. While I did not (and don't want to) grasp all of your explanation I gather that 1 point is useless as you need to select and area of the cloud to work with.
Never underestimate the power of human stupidity
RAH
|
|
|
|
|
OK, you did say that displaying that many points was probably a design problem, right? So how would you suggest designing a system with multi-millions of points and a need to show them to the user so that they can pick 1 or a few? How would you summarize the info and still allow single points to be selected without sacrificing speed? I'm certainly interested in learning something here and if it works, I can apply it immediately.
Mycroft Holmes wrote: I gather that 1 point is useless as you need to select and area of the cloud to work with.
No, I'd prefer to pick just 1 point, and find out where it came from, but the best I've been able to figure out how to do is pick a small area and then let the user refine from there.
CQ de W5ALT
Walt Fair, Jr., P. E.
Comport Computing
Specializing in Technical Engineering Software
|
|
|
|
|
While I don't do a lot of work with charts I do a lot of point analysis. Our people are not interested in points that conform to the rules, they are after the exceptions, we spend most of our time refining the rules so the exceptions are made available to the user for further analysis.
Walt Fair, Jr. wrote: without sacrificing speed
Bloody big machines and some horrendous indexing on the tables. Some of our analysis runs will take 30 min to 3 hours. It is a constant battle to get the best performance from the hardware and database, we are forever tweaking a process, changing an index. We have found that doing basic, regular maintenance on the database can give us excellent benefits. We do not have trained DBA so it is up to us developers to do the best we can.
Never underestimate the power of human stupidity
RAH
|
|
|
|
|
Mycroft Holmes wrote: we spend most of our time refining the rules so the exceptions are made available to the user for further analysis.
Wish that were possible!
Mycroft Holmes wrote: We have found that doing basic, regular maintenance on the database can give us excellent benefits.
Yep, same here. Except the stuff I do is supposed to be interactive. Sometimes we go for a cup of coffee while waiting to interact, though ...
CQ de W5ALT
Walt Fair, Jr., P. E.
Comport Computing
Specializing in Technical Engineering Software
|
|
|
|
|
Two problems: Reduction and summary. Pick your partitioning technique (ie: quadtree for your 2D data) and then summarize distribution/density at node level. Then when you render you can perform appropriate culling for visibility and render summary/points.
Selection from a set of millions of points, and then putting them on the screen, is a fairly common problem in game graphics (and then we draw triangles, texture/light/mangle them up with pixel and vertex shaders and still manage to hammer them out at 60fps) 
|
|
|
|
|
stancrm wrote: It should show about 14.5 million points without performance problem.
First, how many pixels can you monitor show? (1280x1024 = 1.3 million) You simple can't show all those points unless your monitor supports resolutions that no video card can generate.
Second, I don't care which charting library you use, you WILL see a lag graphing 14.5 million points.
|
|
|
|
|
 I know, if I show all those 14.5 million points, I see only a green rectangle..
But my customer want it.. It's hard to explain it..
|
|
|
|
|
Good luck with that. The only thing that has a chance of that kind of performance is DirectX and a good video card. Treat the graph like an object in a video game and you might get what you want.
|
|
|
|
|
I have also tried using directx. About 2-3 millions points is OK, after that you cannot see the lines anymore, you can only see the background. The lines are gone. I think directx has also a problem with that.
|
|
|
|
|
Everything is going to have a problem like that. You're graphing MANY more points than there are pixels on the screen. All you're EVER going to get is a big indecipherable blob on the screen.
|
|
|
|
|
Could you tell me how I can drag a panel on the form?
Thanks in advance.
|
|
|
|
|
Yes I can.
...
Use the mouse drag event and use it to set the location of the panel. If there is not a mouse drag use mouse down combined with mouse move.
Need software developed? Offering C# development all over the United States, ERL GLOBAL, Inc is the only call you will have to make.
Happiness in intelligent people is the rarest thing I know. -- Ernest Hemingway
Most of this sig is for Google, not ego.
|
|
|
|
|
Would you mind explaining the code?
|
|
|
|
|
I have a gridview in which I need to combine two columns in one. One field is an true/false, this column I would like to add the string "Office" if it is true. The other field is a string. The problem I am having is that the field in the grid displays true/false and not the string "Office". I am very new to C# in regards that I am using SQL Server to retieve this data. I am sure that the select statment is correct. But the source on the design page is where my problem is...I think. Can you help?
|
|
|
|
|
I'm assuming your data source for this grid is a DataTable. Perhaps adding a DataColumn with an expression would help.
More Info[^]
|
|
|
|
|
Evening all (depending where you are of course !)
This is going to be tricky to explain but i'll give it my best.
Please bear in mind i am a novice to C#.
I'm going to run a loop to create a 'List' of 10 animals.
I'm using polymorhism to then cast each animal the type i need (cat, dog, etc).
newAnimals[0] = new Cat();
newAnimals[1] = new Cat();
I'm confused as to how i can then give each object a reference name so that i can reference individual object's and invoke methods on them ?
For instance - say i had 5 animas in the list and the first 2 were cats.
I need to be able to say -
cat1.eat();
cat2.play();
Is there a way i can do this ?
Thanks in advance
(loving the forum by the way !!)
|
|
|
|
|
No you can't (I don't think so), except you declare 5 variables like:
cat1 = newAnimals[0];
cat2 = newAnimals[1];
...
Then you can use these variables. However newAnimal[0].eat should work for you...
Life is 5: 3 me, 1 you.
Trying to find the missing part is the meaning of Life.
And sadness is when you find that part!
|
|
|
|
|
If you need to tell which animals are of what type to do something specific with them, then your inheritance tree is probably broken. You can add methods like GetAllCats, tho. Then you can do this:
foreach(Animal animal in newAnimals)
{
Cat c = animal as Cat;
if (c != null)
{
// Add c to a list to return, or do something with c
}
}
The as keyword attempts to cast an object as a derived object, and returns null if it fails.
I would also use a list instead of an array, so you can add as many animals as you like, unless your class requires you to use arrays at this point.
If you just mean accessing the objects, then newAnimals[0], newAnimals[1] is how you reference them.
Christian Graus
Driven to the arms of OSX by Vista.
|
|
|
|
|
Thanks Cristian,
Your reply was v-helpful.
I've since tried
newcat1 = zooAnimals[0] as Cat;
but this doesn't seem to work either.
Can i not cast the object in this way ?
|
|
|
|
|
What does 'does not work' mean ? If zooAnimals is an array of Animals, and Cat is a class derived from Animal, then so long as newcat1 is a Cat and zooAnumals[0] is a Cat, it should work fine.
Christian Graus
Driven to the arms of OSX by Vista.
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
