All error messages are presented in the simulation screen. Some are presented even before a simulation has been run, and some are presented in the simulation log if a simulation fails.
Message | Solution |
---|---|
No materials enabled | Select contaminants |
XXX is empty | Provide time series data |
Value/PDF is missing | Provide parameter value |
You must select at least one contaminant from the context screen.
A time series input lacks data. The time series is highlighted in red in the time series screen.
A parameter is missing either a value (if it is deterministic) or a PDF (if marked as probabilistic). The parameter is highlighted in red in the parameters screen.
Message | Solution |
---|---|
When running probabilistic simulations, the setting Output options should be 'Produce specified output only' | Specify timepoints |
Probabilistic simulation set, but no parameters have been selected for probability assessment | Select parameters |
In a deterministic simulation, the solver will by default report outputs for all time-points that where taken. For a probabilistic simulation the reported time series should be the same for each iteration and the user must therefore specify exactly which time point he/she is interested in. This is done in the simulation settings window in the General tab of the simulation settings window. The output mode must be changed to 'Produce specified output only' and one or more time series must be added.
If no action is taken, the simulation will return data only for the start and end time.
When running a probabilistic simulation values are sampled (randomized) for some parameters using probability density functions. Before the simulation, the user should select which parameters to sample values for - the rest will be assigned their deterministic value. Parameters are selected in the Parameters tab of the probabilistic settings window.
If no parameters are included, the final result will be deterministic - that is, all iterations will be identical.
Message | Explanation | Solution |
---|---|---|
“X is NaN a time Y” | NaN is typically caused when zero is divided by zero. It can also be caused when a function returns a complex number, for instance the square root of a negative number: (-10.5. Read more.. NaN can also be used by the model developer to indicate that a value must be entered for the parameter. | NaN |
“X is infinite at time Y” | In most cases this error occurs as a result of a division of zero. | Infinite |
Could not without reducing | The numerical solver could not find an approximation to the differential equations for a given time step, with the given error tolerances. | Error tolerance |
Simulation never finishes | The progress bar stops for more than 15 minutes | Memory problem,Time out |
MERLIN-Expo will abort a simulation as soon as a NaN is detected in any intermediate calculation result.
NaN is sometimes used as a default value for parameters to assure that users have entered all required values. So, if the NaN is reported for a parameter, assert that its values are correctly set in the parameters screen.
If X is another type of block, you need to study its equation(s) to find the problem. In the information box of the model screen, find the equation for the block. This will show you candidate parameters for the error.
Example:
River.Mass transfer coefficient at the surface water-sediment interface is NaN at t=0
The equation is Dwater·φsed^(4.0/3.0)/(Δsed+Δw·φsed^(4.0/3.0))
It is easy to tell that if φsed is < 0, a NaN would be reported.
NaN is a common error during probabilistic simulations, please read more here.