Don't guess -- read these
instructions!
Everyone who has tried to guess how to do this without bothering
to read the instructions has made serious
errors.
It is much harder to correct a mistake than it is to do it right
the first time.
So, please work through the example below, following the instructions,
before working with real data. You may then want to print these
instructions
and check things off as you do them, especially the first few times.
Most of this page refers to a simple example: a dataset with only one Variable (Grain-Yield) and one Date ('6/26/00), and without any Sort parameter (such as soil depth). The master copies of the database and associated spreadsheets are saved on the Agronomy server, but you can work with copies on another computer, if you like.
1) Open the spreadsheet you want to import and decide which Variables (Grain-Yield, Soil-NO3-N, etc.) need to be imported. For practice...download before.xls by clicking the link, then save the practice file in MyDocs doubleclick MyDocs\before.xls to launch Excel and open practice file.
4) Label columns in the new worksheet: Dataset, Plot, Value.
5) Copy plot numbers into the Plot column as references, so that any corrections to the spreadsheet will be updated automatically. This also eliminates error-prone retyping of numbers. The details will vary among spreadsheets. In our example:
6) Find out what our standard
units are for the each variable to be imported.
Huge errors can occur if you simply assume spreadsheet is already in
our
standard
units. If you can access the LTRAS database on the Agronomy
server,
leave the spreadsheet open while you do the following steps to find
our standard units for your variable.
(Otherwise, email rfdenison@ucdavis.edu to find out about standard
units for your variable and skip to step 4.)
7) Convert to our standard units
and then copy data as references
into
the
Value
column of worksheet Database by a similar process.
Our example already has a column with data in kg/ha, our standard
units.
But to illustrate the process we will use data in lb/acre from another
column.
9) Delete rows without data (caused by blank lines in spreadsheet) in Database sheet by right-clicking each row number and selecting Delete (not Cut or Clear contents).
10) Check your spreadsheet (workbook) before
saving it.
Practice worksheet should look like
this.
If not, ask for help now.
11) Fill out the requested information in the NEWDSET.APR form you opened earlier, including DefaultDate, which is the date of field harvest or sampling, not date of sample analysis. Since you haven't defined Variable in the spreadsheet, you also need to select Grain-Yield as the DefaultVariable if you didn't already do so. Exit Approach. (If you are not connected to the Agronomy server, fill out this same information on a hardcopy of the form.)
If more than one variable was measured at the same time (or on the same sample, whenever the analysis was done), add a column labeled Variable. For example, Soil-NH4-N and Soil-NO3-N are often measured at the same time. It's OK to put these in separate datasets -- Approach will match them based on date and plot number -- but it may be simpler to put them in one dataset. In that case, the Variable column indicates whether the Value in a given row of the Excel 5 spreadsheet is for nitrate or ammonium. Don't make up new names for existing variables!
A Sort column is used if there are data for the same Plot,
Date
and
Variable that should not be averaged together. For
example,
soil data could be sorted into values for 0-15 cm versus 15-30
cm.
The Sort column is also used to indicate weed species, using
standard 5-letter codes.