QUB |
Archaeology and Palaeoecology |
The 14Chrono Centre
Manual for psimpoll and pscomb
Error messages
Contents
Running psimpoll involves an interaction between a user and the software,
and an interaction between a computer and the software. It is the aim in
writing the software to be aware of problems that might arise in either of
these interactions and anticipate them. Where possible, some corrective action
is taken, and the program will continue. In other situations, this is not
possible, and the program will stop with a Program ending message. The
sections below include messages that might arise when the program stops
unexpectedly, and give more detail of why they might have occurred. Other
messages may be given by the operating system, depending on why the failure
occurred, and which operating system is being used. Problems also occur in
other ways, but enabling psimpoll to continue. These may give an explanatory
message on the screen if the program realised something odd was happening, or
else may be evident as unsatisfactory output.
Errors are likely to fall into one or more of the following categories:
- The program could not interpret the input file(s) in some
way: incorrect specification of numbers of taxa or samples,
wrongly ordered, incomplete. This is by far the commonest source
of problems. Failure by user, or failure to communicate what
the program expected. Please check the former first!
- The program did not anticipate some (perfectly legitimate)
combination of options or some type of (perfectly legitimate)
data input. Program incomplete: complain;
- The program failed to do something this documentation
claimed it should do. This may be evident as a crash,
or erroneous output. Complain.
If you encounter an error, whether it produces a crash, or just wrong output,
please follow the following steps:
- Try to replicate the error, to ensure that the problem
was not simply through a typing mistake at the terminal;
- Note carefully when the error appears, by reading messages
that appear on the screen, and taking a look in any output files.
psimpoll is moderately verbose, and this is one of the
uses for all that stuff. Note the message itself, especially
any preceding information on file and line number(see below);
- Check all data input;
- Explore the behaviour of the problem: does it occur with
other inputs? If so, when does it appear, and when not?
- Bring evidence to KDB: include all input files.
Crashes can appear in several ways. Wherever possible, you will get an error
message, usually from the program, occasionally from the operating system.
However, it is also possible that the computer will just `lock up': no
action evident at all. I have made psimpoll verbose to help detect this.
Or other strange things can happen: unexpected rebooting, running of disc
drives, etc. Watch for screen output that includes garbage characters as a
warning of a problem. All these circumstances should be reported to KDB.
If the program does crash, you may need to reboot your computer.
A list of messages that precede an unexpected ending
follows. It is ordered strictly alphabetically, in some cases starting from
after a phrase of the form `ppmain.c 288', which gives information
on the program file and line number that detected the error. To see this, run
psimpoll, but give an erroneous file name. Unless otherwise stated,
any of these messages might occur with either psimpoll or
pscomb. Select the message of interest to get more information.
- agedepth (timescale): failed to converge
on a value for t after 10000 iterations
- asin
- atan
- atan2
- bernshtein: failed to converge on a
value for t after 10000 iterations
- bus error
- CA: residual bigger than tolerance
- calloc: memory allocation error
- Configuration file was created by an earlier
version of psimpoll: there may be inconsistencies
- core dump
- Data set for analyses has zero samples
- Error estimates on ages must be greater
than zero, cannot complete analysis. Check values in C14 file
- exp
- Fatal error detected by operating
system: please report to KDB
SIGABRT: Abnormal termination
- Fatal error detected by operating
system: please report to KDB
SIGFPE: Floating point exception
- Fatal error detected by operating
system: please report to KDB
SIGILL: Invalid function image
- Fatal error detected by operating
system: please report to KDB
SIGSEGV: Invalid access to storage
- Fatal error detected by operating
system: please report to KDB
SIGTERM: Termination request received
- fclose: error in closing a file
- fgetc: no character read
(end of file?)
- fgets: no string read (end of file?)
- First character of two-character code
on line 4 of input must be one of %, A, B, C, G, L, M, or X.
See documentation for details. The character in your file is 'c'
- fopen: failed to open file
- fscanf: unexpected end of file
- getbuf: buffer is overfull. Check that
individual input items do not exceed allowed number of characters.
- getbuf: unexpected end of file.
- getc: reading error (end of file?)
- indsplit: more than the allowed maximum
of 20 splits found. Please report to KDB
- Input box is greater than 63000
characters long
- int conversion: floating point number
cannot be represented as integer
- lfit (gaussj): singular matrix, cannot
complete analysis
- linint_s: two depths are identical or
reversed
- loess: extrapolation beyond date limits
is not allowed. Define an age for top and basal samples in menu L.
- log10
- log
- Memory allocation error. Cannot load
COMMAND, system halted
- Number of samples is zero or negative
- Number of sum taxa is greater than 10
- Number of sum taxa is less than 1
- Number of taxa is zero or negative
- pow
- ran4_s: this function is somewhat
non-portable Output on your computer appears unreliable
Use one of the other random number generators
- Rate of change analysis: the
dissimilarity measure in use has exceeded its theoretical maximum
upper limit
- realloc: attempt to reallocate to a
negative amount of space
- realloc: memory allocation error
- scanf: input failure
- Second character of two-character code
on line 4 of input must be one of A, D, M, S, or T. See
documentation for details. The character in your file is 'c'
- Sediment accumulation rate(s) have been
calculated as negative
This will result in age reversals, and/or
negative pollen accumulation rates. Check output files for
details, and reconsider age model
- Segmentation fault
- SIGINT: interactive attention signal
from user
- spline_s: two depths are identical or
reversed
- splint: bad xa[] input, cannot complete
analysis. Check depths in C14 file
- sqrt
- strcat: total string length too long
- strftime: failed to fill array
- svdfit (svdcmp): no convergence in 30
iterations of its, abnormal exit
- svdfit (svdcmp): zu() needs augmenting
with extra zero rows, abnormal exit
- switch: unexpected value
- SYSTEM:
- TILIA format: line-length unexpectedly
short. Check input file!
- time: calendar time not available
- Unexpected end-of-file reached: data set
is incomplete
- ungetc: error pushing character back
- vprintf/vfprintf: output error
Back to contents page
Copyright © 1995-2007 K.D. Bennett
Archaeology and Palaeoecology | 42 Fitzwilliam St | Belfast BT9 6AX | Northern Ireland | tel +44 28 90 97 5136
Archaeology and Palaeoecology | The 14Chrono Centre | URL http://www.qub.ac.uk/arcpal/ |
WebMaster