- Given: search key
(Age = 50, Salary = 100)
How to find this record:
- Find the row index using
age = 50:
- Find the column index using
salary = 100:
- Find the offset
(this is the standard way to
find an array element):
offset = row index * (column width) + column index = 1 * 3 + 1 = 4
- Access the blocks
and search for the record:
- Find the row index using
age = 50:
- Insert:
insert( record ) { Lookup( record.SearchKey ); Let b = the last bucket block; if ( b has room for record ) { Insert record in block b; } else { Allocate an overflow block for bucket; Link overflow block to b Insert record in overflow bucket block; } }
- Assumption:
- The grid index file can be store in memory
- Lookup performance:
- 0 block access to obtain
the bucket block pointer
- 1 block access to obtain the
data block
(that contains the record)
(If there are overflow blocks, we need to access a few more (overflow) blocks)
- 0 block access to obtain
the bucket block pointer
- Insert performance:
- In addition to
the lookup cost:
- 1 more block write
operation to
update the
bucket block
(If overflow, we need to update the overflow link in the bucket and write an overflow block)
- 1 more block write
operation to
update the
bucket block
- In addition to
the lookup cost:
- Assumption:
- The grid index file can be store in memory
-
Partial Match queries
- The query specifies conditions on some dimensions but not on all dimensions
Example:
- Find all jewelry purchases by
people with
age = 50
Conclussion:
- You will access
m disk blocks
(m is some dimension of the grid)
-
Range queries
- Find objects that are located either partial or wholly within a certain range
Example:
- Find all jewelry purchases
by people whose:
35 ≤ age ≤ 50 50K ≤ salary ≤ 100KSolution:
In this example, we must access 4 disk blocks
-
Nearest neighbor queries:
Find the nearest neighbor of a data point:
The expanding range search algorithm:
- Start by
finding the nearest neighbor
in the
bucket that contains the
data point:
This distance will limit the block where you need to search to all blocks that intersect with this circle:
- Expand the
search region in an
adjacent bucket that
contained
within
the circle:
And so forth:
And so forth:
Note:
- You may need to expand the
search range beyond the
adjacent regions.
Example:
The nearest neight is outside the adjacent regions !!!
- You must use the current nearest neighbor and the grid lines to decide whether you need to expend the range of the search
- Start by
finding the nearest neighbor
in the
bucket that contains the
data point:
- Performance:
- The expanding range search will
access on average
9 data blocks
(in a 2-dimensional grid index):
- The expanding range search will
access on average
9 data blocks
(in a 2-dimensional grid index):
-
Where-am-I queries:
- Given a
location
(i.e., coordinate)
Find the object(s) that contains the location:
- The Grid index
cannot
represent objects
(can only
present points)
- The Grid Index cannot handle Where-am-I type of queries !!!
In fact, the only kind of index that can handle Where-am-I queries is:
- The R-tree (Region-tree)
(Discussed later)
- Given a
location
(i.e., coordinate)