11.02.2013 Views

Full document / COSOC-W-86-002 - the National Sea Grant Library

Full document / COSOC-W-86-002 - the National Sea Grant Library

Full document / COSOC-W-86-002 - the National Sea Grant Library

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Table 2.—Selected SAS Data Management Tools.<br />

Data Input<br />

Data Set Transfer - Among Different Machines<br />

Sorting - Multlplo Sort Fields<br />

Character String Manipulation<br />

Appending of Two or More Data Sets<br />

Merging of Two or Hore Data Sets<br />

Interleaving of Two or More Data Sets<br />

Data Set Subsettlng Based on Logical Testing<br />

Deletion of Variables or Observations<br />

Variable Renaming<br />

General Data Transformations<br />

Maoroa - For Storing Frequently Used Sections of Code<br />

Matrix Processing<br />

Multiple Simultaneous Pile Processing - Input and Output<br />

PROC TABULATE calculates descriptive statistics and displays <strong>the</strong>m<br />

In tables. The tables are suitable for most reports. This procedure<br />

con also calculate frequency counts and percentages. PROC TABULATE<br />

allows users a great deal of flexibility in <strong>the</strong> format of <strong>the</strong>ir<br />

tables. The physioal layout of <strong>the</strong> tables and <strong>the</strong> arrangement of<br />

variables and values in <strong>the</strong> table are controlled by <strong>the</strong> user.<br />

Estuarine Data Quality Aaauranoo Using SAS<br />

Many analyses supporting quality assurance aotivltles can be<br />

carried out using SAS graphical procedures. For example, PROC GPLOT<br />

and PRX GCRART produce scatter plots (with overlaid means and standard<br />

deviations, if desired), bar charts, line graphs, and o<strong>the</strong>r types of<br />

graphical displays. These help <strong>the</strong> estuarine researcher to identify,<br />

e.g., keypunch and o<strong>the</strong>r data processing errors, and discrepancies In<br />

measurement units. In <strong>the</strong> <strong>National</strong> Estuaries Program, data ohooking Is<br />

an Interactive prooess. SAS print-outs are returned to <strong>the</strong> data<br />

submitter or researcher for verification. After oorrootion, tho data<br />

are displayed a seoond time using SAS graphical procedures to double<br />

check <strong>the</strong> correction process.<br />

Ano<strong>the</strong>r Important aspect of estuarine data quality assurance is<br />

<strong>document</strong>ation. The <strong>National</strong> Estuaries Program Data Base Includes many<br />

different types of data oolleotod under a variety of sampling<br />

programs. Field and analytioal procedures mist bo accurately and<br />

oonclsely oommunioated to future Data Base users. Data associated with<br />

unusual circumstances oust be noted. Suoh procedures and notes ore<br />

stored in <strong>the</strong> Data Base through <strong>the</strong> use of SAS variable labels plus<br />

supplemental, accompanying variables. For example, each physical and<br />

chemical variable con be associated with a two byte character variable<br />

In which remark codes are stored suoh as "less than deteotion limit.*<br />

Ano<strong>the</strong>r variable is used to store numeric oodos representing various<br />

analytioal procedures. Examples of SAS variable labels are shown In<br />

Table 1.<br />

443

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!