LoadRunner
Preparing Custom Input Files
Sometimes, one has data that did not come from the USGS. Since making it mimic USGS
data formats can be rather a pain, LoadRunner will accept simpler data input files prepared by
hand.
Note: This is mix and match. Either the flow, or quality, or both, input files can
be hand-crafted.
Quality Data
If you're hand-crafting quality data (qwdata), you need to prepare the data in tab-separated columns. The simplest
possible format is:
# simplest-style qwdata file
#
# Lines beginning with # are comments, and can appear wherever.
# Note, though, that the magic column-decoder line does NOT begin with a comment character.
#
sample_dt p00410
1993-02-08 240
1993-04-01 175.5
1993-04-26 174.5
...
Notes:
-
Required : The tab-delimited magic column-decoder line ("sample_dt(tab)p00410") is required. This
tells LoadRunner which columns provide what data. In the example above, the columns are :
the date in YYYY-MM-DD format, then the concentration of elem code 00410 in the column labeled
'p00410'. This is similar to the USGS qwdata formats, minus many extraneous columns.
- If you don't have a USGS element code for your substance, use p00410. It doesn't really matter - but your
element code XXXXX, of pXXXXX, has to appear on the list you tell LoadRunner to look for.
- The values are tab-delimited.
- You can have "< 0.10" (no tabs, one space) as a column value. It means "below the detection limit of 0.10."
- YYYY-MM-DD : My version of Microsoft Excel doesn't fancy this date format for US English, but it does for
Afrikaans, etc. languages. On import, I told it the column was date of YMD. After import,
I formatted the spreadsheet column into YYYY-MM-DD.
- NoSiteGiven : Because I did not provide a site_no column, this data is assigned a site name
of "NoSiteGiven". Note that unless SplitSite is set to false,
this name has to match the flow data file. (If site_no isn't provided in the flow file either, that site is also "NoSiteGiven".)
Unless you provide a site_no column, you can't have more than one site in your LoadRunner run.
Optional : Both site_no and flow column (p00060) are optional:
Normally, you would provide a site_no column, although it's optional:
#
# Typical hand-crafted qwdata provides site name, date, and concentration value.
#
site_no sample_dt p00410
05030150 1993-02-08 240
05030150 1993-04-01 175.5
05030150 1993-04-26 174.5
...
In this case, I provided a column of site_no, and provided a value for every line. Otherwise, half my data
would be "05030150" and half set to "" for a site name. This site name is a string, not a number. I can name my site
"Xiangdong Province, CN" if I wish (no tabs).
You can also provide a flow value with your concentration value, if you want. Label it "p00060" on the column decoder line:
#
# Hand-crafted qwdata with site names and a flow value.
#
site_no sample_dt p00410 p00060
05030150 1993-02-08 240 98.00
05030150 1993-04-01 175.5
05030150 1993-04-26 174.5
...
In this case, I also provided a column of p00060 - this is streamflow in ft3/sec. Notice that I only provided
a value for this column on one line. That means for that 1993-02-08 only, I insist on a streamflow of 98.0 for calibration purposes,
instead of whatever the
flow file says. (I could also have provided flow values for the whole column if I wanted to.)
The site name and flow columns are independently optional - I could drop either or both.
I could also provide p00061 instead of p00060 for flow. It is also possible to have multiple concentration columns
for your element. Whatever column(s) your concentrations are in, though - say, p00410 and p90410 - must be on the list of elem codes you tell LoadRunner
to look for.
Flow Data
This is the second input file you provide to LoadRunner, and can also be hand-crafted.
Just like the quality data, this is in tab-separated columns, with a magic column decoder line. The simplest
possible format is:
#
# simplest-style flow file
#
datetime 01_00060_00003
1993-02-01 98
1993-02-02 96
1993-02-03 90
1993-02-04 94
...
Notes:
- Take the column decoder line verbatim. They're USGS flow column names. Yes, they don't match the USGS quality file column names.
- NoSiteGiven : See comments above for quality file.
Optional : You can optionally provide a site_no column. The same example with named site:
#
# normal-style flow file includes site name, datetime, and flow columns (named verbatim as below)
#
site_no datetime 01_00060_00003
05030150 1993-02-01 98
05030150 1993-02-02 96
05030150 1993-02-03 90
05030150 1993-02-04 94
...
Ginger Booth for
Peter Raymond,
October, 2007