QUB |
Archaeology and Palaeoecology |
The 14Chrono Centre
Manual for psimpoll and pscomb
Menu
B Scaling of taxa and samples
a Independent scaling for each taxon (off) [off]
b Scaling covers excluded samples (off) [off]
c Use defaults for marks on age/depth scale (on) [on]
d Use two columns for sum values (off) [off]
e Skip every nth sample (0) [0]
f Interpolate to constant sample intervals (off) [off]
g Threshold for missing data values (-0.99) [-0.99]
9 Leave this menu
Allows changes to factors that control various scaling properties:
By default, a scale is found that will work for all taxa in the
dataset, except that for those specifically marked as separate (e.g.,
charcoal, rarefaction
data). This option disables the search for a common scale,
and searches for a separate scale for each taxon;
Toggles from `on' to `off'.
By default, if samples have been excluded, either because they
were marked with an `x' (see Data preparation)
or because they are outside a section of
the sequence that has been selected (menu H),
then data values in those samples
are not considered when scaling plots. This option allows the data values in
excluded samples to be considered. It means that when sections of the diagram
are picked out, each such section will have plots with the same scaling as
the others;
Toggles from `on' to `off'.
By default, psimpoll determines an appropriate scale with
major and minor marks for the age-depth axis. This option enables the user
to turn off the default and specify an interval for both major and minor
marks. Interval must be given in the same units as those in effect for the
age-depth scale (allowing for any conversion to ages). If the default is
turned off, the current settings for major and minor marks are shown, and
then prompts for new values;
Values for pollen sums, identified by marking the taxon `name'
with `*', are normally printed as a single column with upto 50
samples, but as a double column for more than 50 samples. Selecting this
option forces double-column printing however many samples there are;
Toggles from `on' to `off'.
Allows omission of samples, regularly spaced. This is included
to permit exploration of the sensitivity of certain analyses (notably
zonation)
to the number of samples in the dataset. A value of `2' causes
every other sample to be selected for omission, `3' every third sample, etc.
A negative value
causes the selected samples to be included, and the rest omitted: a value of
-3 causes only every third
sample to be included. Thus, `2' and `-2' would both select every other
sample, but the two series would have alternating samples. Values of `1'
and `-1' are not allowed;
[Allowed range -100-100].
Selecting this item produces the following display:
6 Interpolate to constant sample intervals (off) [on]
Enter '0' for on, or '1' for off
Allows a new dataset to be generated, based on the input data set, but
with interpolated samples evenly spaced. The values at the interpolated samples
are based solely on the pair of original samples either side of the
interpolated sample (or on one original sample if there is one with exactly
the same age or depth value as the desired interpolated sample).
If 0 is selected, you will see the following submenu:
A Interpolation interval in age-depth units (1) [100]
B Start level for interpolation (0) [0]
C Stop level for interpolation (0) [10000]
D Interpolation method (nearest pair) [nearest pair]
9 Leave this menu
Q Return to main menu
Here you can select the interpolation interval, using the same units as the
input dataset or age units (if a conversion to age is being done),
specify where the interpolation should start and stop, and determine the
interpolation method. The two available methods are `nearest pair', and
`line fitting'. The first finds a pair of samples that lie either side of
the interpolated level (however far away they are), and interpolates between
them. The second uses all levels upto half of the interpolation interval
either side of the interpolated level, fits a line through these points,
and obtains the desired value for the interpolated level from the line.
For the line-fitting method only, if
there is only 1 (or 0) level available, the value for the interpolated level
is given as for a missing data (-1, by default: see Menu Bg).
If this option is selected, confidence intervals (menu N) are not available, because the value at any level is a combination of values from several levels.
Allows change to the value of the negative number used to identify missing
values. By default, this is -0.99. Anything more negative than this (e.g. -1)
is used to mark a missing value. The value can be changed to any other
negative number. This may, however, cause undesirable results for existing data sets
that use -1, so use with caution.
[Allowed range: any negative real number].
Return to main menu.
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