given data point by its position within the file,
Print option 1 (Voyage summary). Printed output
rather than any identification information in-
for option 1 includes that provided under option 0
cluded in the data record. If the order of data
plus three tables for every voyage that summarize
points is changed, the program may appear to
environmental variables in effect and the estimated
run properly but the results will not be accurate
cost of the voyage in U.S. dollars. The first table
since probabilities will be assigned incorrectly
gives summary statistics (minimum, maximum,
to the wrong data node.
mean, variance, and standard and average devia-
tions) for values assigned to permanent and wind-
Output files and print options
induced currents, waves, winds, environmental
Each time the model is run, simulation re-
factors that slow ship speed (fog, icing, snowstorm,
sults a r e written to an ASCII print file named
and darkness), and ship speed. The second table
NSRSIM1.PRN. If the file already exists, previous
tabulates the length of time and the total distance
results are overwritten; if it does not exist, it
over which different conditions (fog, icing, waves,
is created. To preserve results from a run,
snowstorm, darkness, ice concentration, and ice
NSRSIM1.PRN must either be printed or renamed
thickness) were encountered during the current
before another simulation is run. NSRSIM01 al-
transit. The third table presents the cost of the tran-
lows for three levels of detail in output written to
sit in terms of time and money.
the print file. At the most general level (print
Print option 2 (Listing of variables). Printed out-
option = 0), the program prints only statistics that
put for option 2 includes the tables provided un-
describe the entire transit simulation. At the in-
der options 0 and 1 plus a table that gives a run-
termediate level (print option = 1), the program
ning log of the values assigned to each variable for
prints additional statistics that summarize param-
the entire simulation. A new entry is made in the
eters used to make each individual voyage. At the
table each time any of the simulation variables
most detailed level (print option = 2), the pro-
changes value, and a separate table is constructed
gram prints, in addition to summary statistics
for each voyage.
provided by options 0 and 1, a step-by-step log of
each voyage and lists values assigned to each
Sensitivity analysis mode
variable every time any parameter changes value.
During a typical run, the model assigns values
Appendix C contains examples of the three avail-
to each variable according to MC methods and
able print options.
other techniques embodied in the simulation code.
Print option 0 (Simulation summary). The output
Under some circumstances it becomes advanta-
consists of a series of eight tables. The first table
geous to examine the extent to which transit dura-
summarizes minimum and maximum values ob-
tion and cost depend on a particular variable or
served over all transits for
series of variables. Sensitivity analysis, as this ap-
a) transit time, speed, distance, and cost,
proach is sometimes called, can be applied by set-
b) time with icebreaker escort,
ting the sensitivity variable in the command line
c) time with environmental variables that caused
to 1. When the model is run in sensitivity mode,
reduced speed (fog, icing, snowstorm, darkness,
the program prompts the user with regard to how
waves) and the range of speed reduction factors
values are assigned to each variable. Fixed values
attributed to each variable,
that are held constant throughout the run can be
d) the range of permanent and wind-induced
entered from the keyboard for some variables at
currents, and
the same time that others are selected randomly.
e) hours and distance over which open, ice-free
To run the second example above in sensitivity
ocean was encountered.
mode, change the last digit from 0 to 1, as shown:
Mean values as well as variance and standard
deviation for these parameters are also provided.
NSR 10 3 2 2 5 1
The remaining seven tables give statistics that sum-
marize the time and distance over which different
This will produce output for an October transit of
environmental phenomena were in effect over dif-
a Strekalovsky-class dry bulk carrier traveling east
ferent legs of the transit network. Tables are pro-
to west and will provide detailed output for a
vided for fog, icing, snowstorm, darkness, sea ice,
five-voyage simulation. The short output saves
icebreaker escort, and the number of times each
only the summary statistics for the run. The me-
leg was traversed during the current simulation.
dium output choice saves the statistics for each
34
Previous Page
