Internal architecture of R-tree implementation

Does anyone know where I can get whitepapers which describe the architecture of R-tree implementation in Oracle?
I need it for my undergraduate thesis. It will be a case study of R-tree implementation in Oracle Spatial. Actually I want to focus on why R-tree is implemented while scientists still shape it into a more applicable and reliable state.
But ya, if it is too hard, perhaps I will focus on the architecture itself.
Anyone can help?

Several papers were published in the last few
ICDE, SIGMOD and SSTD conferences about how R-trees are implemented in Oracle.
You can get most of the architecture details from those papers.

Similar Messages

ClusterTools5.0 Internal Architecture

Dear All,
I am a PhD student and are working on HPC area. And I am looking CT5.0 source code to understand the special feature of it different with other MPI implementation. But it has very few documentation, unlike others.
Does any one suggest some books, papers, sldes or website for Cluster Tools internal architecture?
Thanks very much
Noah Yan

Online marketing is not PIA. It is a separate web app. The standard PIA PeopleTools concepts do not apply to online marketing.
1. You determine what happens when a user clicks a button in online marketing through the online marketing dialog designer. You can execute PeopleCode or Java at any point in the dialog by adding a custom action. You can find an example of a custom Java action and a custom PeopleCode action in the PeopleBook "PeopleSoft Enterprise Online Marketing 9 PeopleBook > Using Extensions."
2. Yes, you can manually create profile tables for OLM. Sorry, I can't remember how to do this. I have never needed to do this. It appears that you know how to create a profile and use those profile fields in dialog designer. If you are interested in looking at the profile field data and responses, then look at PS_RA_PROFILE, PS_RA_ATTRIBUTE, and PS_RA_ATTR_CHOICES. If you want to display this data on a PIA page or report off this data, then you can create views against these tables.
3. I believe that OLM (online marketing) just stores choices/answers in the RA tables. I don't think it updates any student admin tables. I could be wrong on this though. Your rep should be able to put you in contact with someone that can give you a definitive answer on that. If you have to update the student admin tables yourself, then I would create the profile fields with whatever names seem appropriate to you, then use a PeopleCode custom action to either insert the prospect's answers into the student admin tables or run the answers through a CI (CI is preferred).
CRM and Siebel are different products. CRM is the PeopleSoft product acquired by Oracle through the PeopleSoft acquisition. CRM is built using the PeopleTools technology. Siebel was acquired by Oracle through a separate acquisition and is a CRM product built using a different toolset.

A Binary Tree Implementation in ABAP

Hi,
Can any one explaine me how to create a binary tree of random numbers with dynamic objects.
Thanks,
Manjula.

Hi manjula,
This sample code uses dynamic objects to create a binary tree of random numbers as per your requirement ...pls go through It.
It stores numbers on the left node or right node depending on the value comparison with the current value. There are two recursive subrotines used for the building of the tree and printing through the tree.
For comparison purpose, the same random numbers are stored and sorted in an internal table and printed.
*& Report YBINTREE - Build/Print Binary Tree of numbers *
report ybintree .
types: begin of stree,
value type i,
left type ref to data,
right type ref to data,
end of stree.
data: tree type stree.
data: int type i.
data: begin of rnd occurs 0,
num type i,
end of rnd.
start-of-selection.
do 100 times.
generate random number between 0 and 100
call function 'RANDOM_I4'
exporting
rnd_min = 0
rnd_max = 100
importing
rnd_value = int.
store numbers
rnd-num = int.
append rnd.
build binary tree of random numbers
perform add_value using tree int.
enddo.
stored numbers are sorted for comparison
sort rnd by num.
print sorted random numbers
write: / 'Sorted Numbers'.
write: / '=============='.
skip.
loop at rnd.
write: rnd-num.
endloop.
skip.
print binary tree. This should give the same result
as the one listed from the internal table
write: / 'Binary Tree List'.
write: / '================'.
skip.
perform print_value using tree.
skip.
*& Form add_value
text - Build tree with value provided
-->TREE text
-->VAL text
form add_value using tree type stree val type i.
field-symbols: <ltree> type any.
data: work type stree.
if tree is initial. "When node has no values
tree-value = val. " assign value
clear: tree-left, tree-right.
create data tree-left type stree. "Create an empty node for left
create data tree-right type stree. "create an empty node for right
else.
if val le tree-value. "if number is less than or equal
assign tree-left->* to <ltree>. "assign the left node to fs
call add_value recursively with left node
perform add_value using <ltree> val.
else. "if number is greater
assign tree-right->* to <ltree>. "assign the right node to fs
call add_value recursively with right node
perform add_value using <ltree> val.
endif.
endif.
endform. "add_value
*& Form print_value
text - traverse tree from left-mid-right order
automatically this will be sorted list
-->TREE text
form print_value using tree type stree.
field-symbols: <ltree> type any.
if tree is initial. "node is empty
else. "non-empty node
assign tree-left->* to <ltree>. "left node
perform print_value using <ltree>. "print left
write: tree-value. "print the current value
assign tree-right->* to <ltree>. "right node
perform print_value using <ltree>. "print right
endif.
endform. "print_value
pls reward if helps,
regards.

Internal error in R-tree processing: [Recursive fetch error]

Hello,
I seem to be getting an error when using any type of SDO function (SDO_RELATE, SDO_FILTER, SDO_ NN, ... ) on my spatially indexed data. The error I'm getting is this:
ERROR at line 1:
ORA-29903: error in executing ODCIIndexFetch() routine
ORA-13236: internal error in R-tree processing: []
ORA-13236: internal error in R-tree processing: [Recursive fetch error]
The interesting part of this is that I only get this error if I insert more than 31 entries into the table. All the relevant code is posted below, what I do is create the table, then create the metadata from the sqlplus interface. After that I run a java program that does Clear() then Insert() then Create_Indices(). All these run just fine (at least run without telling me about any errors). But as soon as that is all done and I do an SDO query on the data, for example:
SELECT location stop_range_poly_area
FROM stops
WHERE SDO_FILTER
location,
SDO_GEOMETRY
2003, null, null,
SDO_ELEM_INFO_ARRAY(1,1003,1),
SDO_ORDINATE_ARRAY(300,300, 600,300, 600,600, 300,600, 300,300)
) = 'TRUE';
I get the above mentioned error. But if I change my data so that it doesn't have more than 31 entries (doesn't matter which 31 entries, just so long as it doesn't exceed that number) that query (and all of my other SDO test queries on the stops table) seems to work just fine.
Thanks for looking at this,
Brad
---- SQLPLUS Code -----------------------------------------------------------------------------------------------------------
CREATE TABLE stops
stop_id VARCHAR2(4) PRIMARY KEY,
address VARCHAR2(256),
location sdo_geometry
INSERT INTO user_sdo_geom_metadata VALUES
'stops',
'location',
MDSYS.SDO_DIM_ARRAY
MDSYS.SDO_DIM_ELEMENT('X', 0, 600, 0.005),
MDSYS.SDO_DIM_ELEMENT('Y', 0, 800, 0.005)
null
----- Java Code -----------------------------------------------------------------------------------------------------------------
// Clear
// Description:
// This function clears all the data in the stop tables in the database
// given in the connection.
public static void Clear (Connection connection)
Statement statement;
try
statement = connection.createStatement();
statement.executeUpdate("DROP INDEX stops_index");
statement.executeUpdate("DELETE stops");
statement.close();
catch (SQLException e)
System.err.println("SQLExcpetion (Stop.Clear()): " + e.getMessage());
// Create_Indices
// Description:
// Create indices for stops table
public static void Create_Indices (Connection connection)
Statement statement;
String sql_query;
sql_query = "CREATE INDEX stops_index ON stops(location) " +
"INDEXTYPE IS MDSYS.SPATIAL_INDEX";
try
statement = connection.createStatement();
statement.executeUpdate(sql_query);
statement.close();
catch (SQLException e)
System.err.println("SQLExcpetion (Stop.Create_Indices()): " + e.getMessage());
// Insert
// Description:
// This function inserts this object into the stops table in the database
// given in the connection.
public void Insert (Connection connection)
Statement statement;
String sql_query;
sql_query = "INSERT INTO stops VALUES ('" +
stop_id + "', '" +
address + "', " +
"sdo_geometry(2001, null, sdo_point_type(" +
location.x + "," +
location.y + ",null), null, null)" +
try
statement = connection.createStatement();
statement.executeUpdate(sql_query);
statement.close();
catch (SQLException e)
System.err.println("SQLException (Stop.Insert()):" + e.getMessage());
Message was edited by: loos to include the changes proposed by the second poster.

Hi,
Thanks for trying but changing those items for the specific failing queries didn't seem to help. Though you do seem to be right about the internal/external polygon problem, so I changed it all my other queries to see if they would fail (that way things would at least be consistent), but it doesn't seem to make a difference. All my old failing quries still fail and my working ones still work. I also changed the co-ordinates as you specified and still have no changes (unless of course the results changed, but right now I'm just looking for queries that compile and run, the results don't matter yet).
Maybe this will help, I'll give you guys both sets of data, one that works and one that doesn't and maybe you can see a problem in the data that I'm just missing or too ignorant to see. The data is simply comma seperated values that I parse into the required fields in the order (id, description, x, y).
So far, the only reason I've been able to find that the non-working data doesn't work is because there are more than 31 rows. I started taking records out of the stops table in a binary search sort of pattern. If I take out stops 100-115 (resulting 29 records) all the queries work, if I take out stops 100-107 (resulting in 36 records) it doesn't work. If I take out 109-115 (resulting in 35 records) it doesn't work. If I take out 1-11 (resulting in 32 records) it doesn't work. If I take out 1-12 (resulting in 31 records) it does work. Here's a table
Take out Stops ---- records left -- Works?
s1-s11 ------------ 32 ------------ No
s1-s12 ------------ 31 ------------ Yes
s100-s115 --------- 29 ------------ Yes
s100-s107 --------- 36 ------------ No
s109-s115 --------- 35 ------------ No
Thanks again for checking this out,
Brad
------- Working Data ---------------------------------------------------------------------------------------
(Student_id, Department, x, y)
Student_1,Computer Science ,296,131
Student_2,Social Science,130 ,279
Student_3,Mechanical Engineering ,392,180
Student_4,Electrical Engineering ,342,322
Student_5,Computer Science ,165,490
Student_6,Scicology ,393,533
Student_7,Physical Therapy ,590,616
Student_8,Civil Engineering ,165,640
Student_9,English ,360,412
Student_10,Economy ,89,32
Student_11,Computer Science ,26,117
Student_12,Social Science,430 ,291
Student_13,Mechanical Engineering ,382,80
Student_14,Electrical Engineering ,542,222
Student_15,Computer Science ,154,290
Student_16,Scicology ,493,323
Student_17,Physical Therapy ,290,426
Student_18,Civil Engineering ,65,230
Student_19,English ,300,412
Student_20,Economy ,44,292
Student_21,Computer Science ,146,431
Student_22,Social Science,405 ,179
Student_23,Mechanical Engineering ,192,480
Student_24,Electrical Engineering ,412,202
Student_25,Computer Science ,265,49
Student_26,Scicology ,33,273
Student_27,Physical Therapy ,186,216
Student_28,Civil Engineering ,365,600
Student_29,English ,309,42
Student_30,Economy ,415,392
------- Non Working Data ---------------------------------------------------------------------------------
(Stop_id, Address, x, y)
s1, 2341 Portland,377,64
s2, 24th St. / Hoover St.,308,22
s3, 2620 Monmouth Ave.,272,138
s4, 2632 Ellendale Pl.,128,110
s5, 2726 Menlo Ave.,85,231
s6, 2758 Menlo Ave.,84,124
s7, 28th St. / Orchard Ave.,183,236
s8, 28th St. / University Ave.,414,308
s9, 30th St. / University Ave.,391,352
s11, 34th St. / McClintock St.,180,458
s12, 36th Pl. / Watt Way,176,622
s13, Adams Blvd. / Magnolia Ave.,218,87
s14, BG Mills Apts.,23,637
s15, Cardinal Gardens Apts.,156,389
s16, Centennial Apts.,373,126
s17, Chez Ronee Apts.,446,414
s18, City Park Apts.,70,323
s19, Dental School,219,478
s96, Founders Apts.,373,192
s97, Hillview Apts.,412,214
s98, House of Public Life,531,303
s99, JEP,304,523
s100, Kerchoff Apts.,473,272
s101, Leavey Library,370,559
s103, McClintock St. / Childs Way,129,553
s104, Mt. St. Marys College,565,127
s105, Pacific Apts.,398,240
s107, Parking Center,525,652
s109, Parkside,78,651
s110, Severance St. / Adams Blvd.,435,202
s111, Research Annex,492,776
s112, Sierra Apts.,352,230
s113, Sunset Apts.,267,278
s114, Terrace Apts.,156,280
s115, Troy East Apts.,402,397
s116, University Regents Apts.,182,181
s117, Watt Way / 36th Pl.,176,622
s119, Watt Way / Bloom Walk,158,653
s120, Windsor Apts.,257,236
s121, Zemeckis Center,476,474
s137, Gate #2,321,715
s138, 24th St. / Toberman St.,377,64

ADF BC+ Faces Internal architecture

Can someone share me where i can find the Internal architecture of ADF BC+ Faces.
How Entity VO, App Module ,Backing/Managed Beans interact.
Thanks in Advance.

Thank you very much Frank/Shay/Timo..
Iam looking for sequence diagrams for a typical request life cycle(from the jspx to the entity object) ..
Here are some querys lingering in my mind..
1) In typical Faces environments (JSF) When a form is submitted the data submitted (anything..even the Value objects) will be available in the Managed Bean where we can play with the data (thats being submitted).
How can i access the Value Object getting submitted in ADF the same way ?? That increases our flexibility
2) I visualised the AppModule is a kind of facade, and it plays around with the ORM entity objects to get the CRUD jobs done..Is this correct??
From the entity object generate code it seems its getting injected somewhere..(dependency injection)
So when a user submits a button.. the request goes to the Managed Bean..calls the App Module.. which calls the Entity Objects Crud operation..
(Now WHere does Updateable VO fit in.. )
Thanks in advance
Edited by: user11922045 on Oct 27, 2010 9:03 AM

Badi:BADI_SD_SALES(SAP says it is for internal use): But i was implemented

Hi Experts,
Badi:BADI_SD_SALES(SAP says it is for internal use): But i was implemented this badi. It is legal or not.
I was implemented badi: BADI_SD_SALES . But it is for SAP internal use.
For implementing this i was changed the field u2018internalu2019 in Standard table: SXS_ATTR.
I removed the X in that field.
All these things done for Creating Return order with reference to invoice. The client requirement is to was external number range instead of internal number rangees.
For this thing I was implement the method t u2018TRANSACTION_INITu2019. I passed the
Our external number in that vbak-vbeln..
It works correctly in development. I am not yet tested in quality and production.
This way is correct one or not. When I removed the u2018Xu2019(Internal use) in field:
SXS_ATTR. In future it gives any other problems or it is legally correct or not.
Helpful answers definitely rewarded.
Regards
Pratap.M

Hello Pratap
Labelling this BAdI as "Internal Use only" SAP can do whatever they want with this BAdI - they may even remove it !!!
Thus, your development relies on something very unreliable.
Perhaps SAP releases this BAdI for general use in the future - or not.
Therefore, I would recommend to use the enhancement framework instead if already available in your SAP system.
Final remark: If you set your implementation of this internal BAdI productive and the customer runs into trouble later on when SAP changes or deletes this BAdI with support packages YOU are liable - and not SAP.
Regards
Uwe

Oracle database Internal architecture

Hi
Any one Can explane me about oracle database internal architecture
thanks & regards

user9098698 wrote:
Hi
Any one Can explane me about oracle database internal architectureYes, there are people who can explain this to you.
How much time do you want to spend? I've been learning the Oracle internal architecture for over 20 years and there are still new things to learn. (Well ... it also depends on what you mean by 'internal architecture'.)
If you want the basics - the Oracle Documentation set at http://tahiti.oracle.com has a Concepts manual that does a decent job of describing the basic architecture. And the Windows Platform Guide, the Unix Administrator's Reference give a fair bit of additional information. As does the Oracle Networking reference.
And there are courses. Oracle University courses provide a decent overview of the architecture. And Jonathan Lewis, Richard Foote and others give excellent courses and seminars on specific aspects of the architecture.
ANd there are books. Search for books by Tom Kyte. Robert Stackowiak.
And internet search.
And http://orainfo.com/architecture/architecture.htm
And ...

AVL tree implementation

Hi all,
I am interested in an AVL tree implementation in Java. I prefer it to be Open Source. Efficiency is my main concern, although a readable code would be appreciated.
The question is if it exist a standard implementation or an implementation you know it is stable and efficient.
thank you in advance,

prometheuzz wrote:
The TreeSet and TreeMap implementations from the java.util package are backed up by a red-black tree. So those classes have the same characteristics as an AVL tree: they're self balancing, and insertions and lookups are O(log(n)) where n is the total number of nodes in your tree.
So, as far as I see, you can use those classes. If not, please explain why you need an AVL tree instead of a proven/reliable class already present in the standard API.Thank you for the answer.
to be honest, in the past, I rejected the idea of using TreeMap or TreeSet, but I fail to understand (or remember) why.
The only reason I see at the moment is that I would like the tree notion of child nodes etc, although I can live without that given that I have an efficient implementation in zero-time. :)
The use of AVL trees is not needed in the implementation, it is just an offspring of the theoretical complexity analysis of what I am about to implement.

ExtJS Drag and Drop Tree Implementation

Hello!
I found this wonderful example of a drag and drop tree, with the ability to save the changes, but I'm not quite sure how to go about implementing it. Can someone point me in the right direction?
Thank you,
Tammy

Hi Tammy
My example uses drag and drop also, so the details are there.
Regarding your question on how to get started:
Really it's just a case of breaking the problem down into managable parts until you have solved all the pieces.
Start by implementing the "AJAX form using globals" at http://apex.oracle.com/pls/otn/f?p=200801:2011
It shows you how to pass data back to the database using AJAX with the WWV_FLOW variables, which are there for exactly that purpose.
Make sure you also understand the response Ext is expecting, use Firebug to examine the responses in both examples.
Once your comfortable with that, try loading a tree using an application process.
You can create stubbed responses initially for the application process, e.g.
htp.p(^'[{
    "text": "adapter",
    "id": "source\/adapter",
    "cls": "folder"
    "text": "dd",
    "id": "source\/dd",
    "cls": "folder"
    "text": "debug.js",
    "id": "source\/debug.js",
    "leaf": true,
    "cls": "file"
}]^');From there, add drag and drop calls, replace your stubs with production code...
I use application processes as a gateway, simply passing requests through to a database package.
It's not a small task by any means, so allow plenty of time for learning and understanding.
Regards
Mark

Binary Tree Implementations - why both key and value?

Hi there!
I'm sitting here implementing a binary tree, and I was just wondering why most of the binary tree examples encapsulate both a key AND a value in their nodes. Wouldn't it be sufficient to just store comparable objects in the nodes (or provide a comparator for alternative ordering)?
Thanks again, Oliver
P.S.:Any suggestions on good books (free E-Books preferred :-)) on ADTs and algorithms? I'm reading 'Data Structures And Algorithms In Java' by Robert Lafore and not very happy with it (the examples are just plain awful).

Trollhorn wrote:
Hi there!
I'm sitting here implementing a binary tree, and I was just wondering why most of the binary tree examples encapsulate both a key AND a value in their nodes. Wouldn't it be sufficient to just store comparable objects in the nodes (or provide a comparator for alternative ordering)?Yes.
Thanks again, Oliver
P.S.:Any suggestions on good books (free E-Books preferred :-)) on ADTs and algorithms? I'm reading 'Data Structures And Algorithms In Java' by Robert Lafore and not very happy with it (the examples are just plain awful).As online resources:
[http://www.cs.princeton.edu/introcs/44st/]
[http://ocw.mit.edu/OcwWeb/Electrical-Engineering-and-Computer-Science/6-006Spring-2008/CourseHome/index.htm]
I don't know of any eBooks on the subject, but I can recommend an excellent hard copy book:
Introduction To Algorithms, by Cormen et al. [http://highered.mcgraw-hill.com/sites/0070131511/information_center_view0/]

Tree Implementation - Like a file system

Hi all,
I'm want to implement something that will look similar to Windows's file explorer - it should have a tree on the left side and a table (datagrid) on the right side. The tree displays some hierarchy, let's say folders, and upon selection on a folder in the tree, the right side displays the files inside it. Important to say that the tree doesn't show leafs (files), just like in the file explorer in Windows.
The entire hierarchy is supplied as an XML to my flex application.
My XML structure will look something like this:
<FOLDER label="folder1" id="folder1_id" >
    <FILE label="FILE1.1" id="FILE11_id" />
    <FILE label="FILE1.2" id="FILE12_id" />
</FOLDER>
<FOLDER label="folder2" id="folder2_id" >
    <FILE label="FILE2.1" id="FILE21_id" />
    <FILE label="FILE2.2" id="FILE22_id" />
</FOLDER>
How can I hide the leafs (the "File" tags) in the tree but show them on the table when clicking on the parent folder? How can I "slice" the XML and supply the children (and only them) to the table (datagrid)?
Thanks!

Yes. This possible with ADF Faces RC, which provides an af:tree component. Have a look at the Fusion Order Demo, which uses the tree to implement a similar usecase to the one you describe.
http://www.oracle.com/technology/products/jdev/samples/fod/index.html
Regards,
RiC

R-Tree Implementation in d-dimension

Hi
I'd like to know if exists a Java implementation fo R-Tree. I founded something in 2-dimensions, but I need one in d-dimensions.
Thank you.

hi,
take a look at
[http://sourceforge.net/projects/spatialindexlib/|http://sourceforge.net/projects/spatialindexlib/]
Then there is a Lucene which is full-featured text search engine written entirely in Java [http://lucene.apache.org/|http://lucene.apache.org/]
Regards,
Alan Mehio
London, UK
Edited by: alan_mehio on Apr 5, 2008 11:00 PM
Edited by: alan_mehio on Apr 5, 2008 11:02 PM

Tree implementation in webdynpro project

Hi,
i want to devlop an application that shows a tree that gets data fron RFC .
user will logged from portal & application will show all his/her subordinates in tree & on click on a
subordinate, all other levels get displayed(if level are in system).
plz help.
Regards,
Amit

Hi Amit,
If the hierarchy of the tree is know at design then you can implement the NonRecursive tree. Refer to the help.sap.com on this:
http://help.sap.com/saphelp_nw70/helpdata/EN/c2/abfceffbacd24185ec7af8a3a2e76a/frameset.htm
However if the hierarchy is not known, then you will have to implement the Recursive Tree. Please refer to the following link in help.sap.com for info:
http://help.sap.com/saphelp_nw70/helpdata/EN/16/1ec1814e566f4baf943c53ccf48552/frameset.htm
The above links also have code to explain how the tree can be populated.
Populate the tree to the necessary level on initialization of the screen or the required action. The remaining tree can be filled when the user clicks on the nodes.
If the child nodes/items are to be populated on click of the parent node, then you can use the onLoadChildren action. When a node is click the action linked to the onLoadChildren is called and the necessary processing can be done to load the data if present.
Regards,
Kartikaye

Binary Tree Implementation

Hi,
How to implement Binary search tree in java?
How to create node & how to do backtracking?
Thanks & Regards,
Sathyavathy S

Pretty much like in most other OO languages.
What exactly is your problem? Do you have a bug in your code, a compiler error or don't you understand how a binary tree works?

Org.jcp.xml.dsig.internal.dom.DOMKeyInfo equals() method implementation

I use something like this:
if(myKeyInfo.equals(anotherKeyInfo))
//I believe two signers are same. It seems correct.
Behind the scene the KeyInfo is implemented by org.jcp.xml.dsig.internal.dom.DOMKeyInfo.
I would like to have a look at org.jcp.xml.dsig.internal.dom.DOMKeyInfo to find out how its equals method is implemented.
Basically I would like to make sure wether myKeyInfo.equals(anotherKeyInfo) is the right way to ascertain the signer.

You can find the source in a few different places, either in the Apache XML Security project repository or the JDK 6 and OpenJDK projects on java.net. Here is a pointer to the JDK 6 source:
https://jdk-jrl-sources.dev.java.net/source/browse/jdk-jrl-sources/jdk6/trunk/j2se/src/share/classes/org/jcp/xml/dsig/internal/dom/DOMKeyInfo.java?rev=3&view=log

Internal architecture of R-tree implementation

Similar Messages

Maybe you are looking for