X射线-物相XRD-PDF4学习资料-23：The Need for Speed

The

Need

for

SpeedThe

PDF‐4+

database

is

designed

to

handle

very

large

amouof

data

and

provide

the

user

with

an

ability

to

performextensive

data

mining.The

database

also

incorporates

many

sophisticated

algoricalculate,

analyze

and

display

diffraction

data.By

design,

speed

is

sacrificed

whenever

there

is

a

trade‐obetween

speed

and

capability.

However

speedimprovements

are

made

annually

and

this

tutorial

gives

mhelpful

hints

on

how

to

improve

speed

for

anydatabase

product.Speed

Gobblers(Speed

bumps)Sophisticated

on‐the

fly

calculatInadequate

computer

hardware

Sorts

and

searches

producing

orusing

very

large

tablesSpeed

Accelerators(Turbo)Smaller

data

sets

Targeted

and

selective

(smaller)searproducing

small

tablesSpeed

BasicsThe

processorand

system

memory

control

the

overall

speed

of

the

databaseand

searches

that

use

the

database.

If

you

exceed

the

specifications

the

systework

faster,

if

you

do

not

meet

the

specifications

certain

operations

will

worslowly

or

notwork

at

all.Search

SpeedsThere

are

several

search

speeds

given

in

this

presentation.These

were

given

to

present

the

reader

with

a

concept

of

relative

spfor

several

types

of

searches.Your

search

times

may

vary

depending

on

the

exact

computer

specificaof

your

PC.The

PC

used

in

these

tests

was32‐bit

with

a

Vista™

operating

system2

GB

RAM

and

2.00

GHz

dual

core

processor.How

fast

is

this

spinning

?The

diffraction

patterns

should

besmooth

and

continuously

drawn.The

shark

should

swimin

a

full

circle

in

underone

second.Speed

Basics

All

Release

2006

and

later

PDF‐2

and

PDF‐databases

use

Sybase

as

a

database

platformand

use

JAVA

as

a

software

platform.

iAnywhere,

the

producer

of

Sybase,

ICDD,

aJAVA

all

continuously

upgrade

their

systemimprove

search

speeds.

You

should

see

fastespeeds

with

each

product

release,

if

youdocomparative

analysis.Database

Search

OptionsSearch

OptionsRelease

2009PDF‐4+PDF‐253

Searches,

291,440

Data

Set49

Searches,

218,610

Data

SetSearchPDF‐4/Organics48

Searches,

370,844

Data

SetSearches

can

be

combined.DisplayDisplay

OptionsPDF‐4+

85

Display

FieldsPDF‐2

24

Display

FieldsDefault

page

displays

8

fields.

This

can

bereduced

to

1

or

expanded

to

all

fields.Search

Speed

–

Size

of

the

SearchThe

larger

the

search,

the

slower

the

speed.Display

all

fields

(85)

on

all

Inorganic

materials[

85

X

262,365

=

22,301,025

Fields

in

the

table]Search

takes

a

minuteDisplay

default

fields

(8)

onall

Inorganic

Materials[

8

x

262,365

=

2,098,920

Fields

in

the

table]Search

takes

~

23.8

SecondsDisplay

PDF#,

chemical

and

mineral

name

for

allZeolites[

3

x

3,155

=

9,465

Fields

in

the

table

]Search

takes

~

Blink

of

the

eye

(0.4

Seconds)HistoryThe

history

panel

tracks

the

search

speeSearches

3

to

5,

displayed

3

fields

for

each

entry

so

the

time

is

directly

relatto

the

number

of

entries

“hit”.

Search

6

at

94.4

secondswas

identical

to

sea5

(8.8

seconds),

except

that44fields

were

displayed

instead

of

3.No.

of

entriesSearchspeedLarge

Calculations

Slow

Things

DoIntegral

IndexIntegral

Index

–

does

a

point

by

point

comparison

of

imported

diffraction

pin

comparison

to

dynamically

generated

experimental

and

calculated

pattern(see

tutorial

on

integral

index).Each

pattern

is

thousands

of

points,

so

if

large

collections

of

patterns

are

uthe

calculation,

then

the

integral

index

calculation

takes

time.Digital

Pattern

CalculationsFor

digital

patterns,

three

algorithms

are

used

depending

on

the

type

of

dataavailable

in

the

reference

material.

The

most

resource‐intensive

algorithcalculations

of

patterns

from

atomic

coordinates.

Calculation

of

a

single

pis

done

in

less

than

a

second.

Calculation

of

thousands

of

patterns

takes

minuResults

FormIn

this

example,

six

entries

were

highlightedand

a

“click”the

right

hand

button

of

the

mouse

produces

all

six

patterns,nearly

instantaneously.Digital

PatternsThis

calculation

of

77

superimposed

explosives

patterns

took

sevseconds.The

more

entries

in

the

calculation,

the

longer

it

takes.Capability

Trade

OffsDatabase

searches

are

frequently

used

for

data

mining.

Manydata

mining

examples

are

provided

in

the

tutorials.A

preferred

technique

in

data

mining

is

to

use

broad

search

parameters,analyze

the

results

and

then

apply

more

restrictions

as

you

find

materialsof

interest.

For

example,

a

researcher

might

analyze

all

zirconia‐contamaterials,

then

focus

on

yttria

or

ceria

stabilized

zirconia

and

then

reduthe

candidate

list

to

tetragonally

stabilized

zirconias,

if

they

werestudying

cutting

tool

compositions.To

be

effective

in

the

trade‐off

between

capability

and

speed,

one

mightwant

to

do

computationally

intensive

calculations

(i.e.

pattern

calculatIntegral

index)

in

the

later

stages

of

data

mining,

when

the

candidate

listhas

been

narrowed.Preference

Options

inSIeve+SIeve+This

is

the

preferencesTable

for

SIeve+.There

are

several

optionsthat

can

increase

or

decrethe

search

speed.Pattern

GOM

and

IntegralIndex

both

involve

time‐consuming

calculations.To

increase

speed,Toggle

these

off

by“point

and

click”.Search

Window0.15Match

Window0.15GOM

Limit2000No

of

Candidates5,885Time5

sec0.150.154000804<

1

sec0.060.06400010<<1

sec0.060.0610001,136<

1sec0.180.18100011,4505

sec0.180.0610004,1761

secSearch

and

Match

Windowsand

GOM

(Goodness

of

Match)The

wider

the

search

and

match

windows,

the

morecandidates

are

reviewed

andselected.

The

moredata

being

compiled

for

display,

the

slower

the

speed.This