99 106 Histogram to estimate result set of σ A cR If a histogram is available

# 99 106 histogram to estimate result set of σ a cr if

• Notes
• 106

This preview shows page 99 - 106 out of 106 pages.

99 / 106
Histogram to estimate result set of σ A = c (R) If a histogram is available for the attribute A , the number of tuples can be estimated with more accuracy. The range in which the value c belongs is first located in the histogram . |B| : number of values per bucket (# distinct values appearing in that range) #B : number of records in bucket T ( σ A = c (R) ) = # B | B | 100 / 106
Histogram to estimate result set of σ A = c (R) Example R(A,B,C) is a relation. T(R) = 10,000 V(R,A) = 50 Estimate T ( σ A =10 (R) ) The DBMS has collected the following equi-width histogram on A range [1,10] [11,20] [21,30] [31,40] [41,50] tuples in range 50 2000 2000 3000 2950 T ( σ A =10 (R) ) = # B | B | = 50 10 = 5 101 / 106
Join Size using Histograms R 1 S Use: T(R 1 S) = T(R) × T(S) max ( V(R,A),V(S,A) ) Apply for each bucket 102 / 106
Join Size using Histograms V(R,A) = V(R,A) = bucket size |B| T(R 1 S) = buckets #B(R) × #B(S) |B| 103 / 106
Advanced Techniques Wavelets Approximate Histograms Sampling Techniques Compressed Histograms 104 / 106
Summary As should be clear by now, result size estimation is not an exact art To estimate the size of the intermediate relations, we have used parameters like T(R) and V(R,A) The DBMS keeps statistics from previous operations to be able to provide such parameters However, computing statistics are expensive and should be recomputed periodically only: statistics usually have few changes over a short time even inaccurate statistics are useful statistics recomputation might be triggered after some period of time or after some number of updates 105 / 106
Outline Estimating cost of query plan Estimating size of results Estimating # of IOs (next) Operator Implementations Generate and compare plans 106 / 106

#### You've reached the end of your free preview.

Want to read all 106 pages?

• Fall '19
• Joseph Rosen
• Relational model, StarsIn, R 1c S

### What students are saying

• As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

Kiran Temple University Fox School of Business ‘17, Course Hero Intern

• I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

Dana University of Pennsylvania ‘17, Course Hero Intern

• The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

Jill Tulane University ‘16, Course Hero Intern