| |
|
IDADE2
Estimating age at death of skeletal individuals
|
Required Inputs
IDADE2 has been optimized to predict age at death of skeletal specimens.
IDADE2 requires an input:
- File description of reference specimens:
IDADE2 reads from a file descriptions of the specimens in the
reference collection. For each specimen, these descriptions include
known age at death of the specimen, and the states of the variables
observed for that specimen. Each specimen is represented by a single
line with the known age at death, the descriptive states, and any
other data, all separated by tabs or commas and in the same order, as in the
*.CSV (comma separated variables) format of EXCEL. In fact,
observations of the specimens in the reference collection may be
organized in EXCEL and 'saved' in *.CSV format to be read directly
by IDADE2.
The specific format of this file is as follows. The first row contains these headings:
- ind: individual identification.
- age: known age of death of the individual. This column
must have the name "age".
- V1-V7: values of the variables.
If you paste the information, these fields should be separated by a
comma, tab or semicolon. If you were using an EXCEL spread sheet to organize
your data, then these headings would be in the first row.
In the example to follow below, data are separated by tabs
ind,age,V1,V2,V3,V4,V5,V6,V7
8r1-X,54,1,3,3,2,4,4,3
18,49,1,3,3,2,4,3,3
C-28,41,2,3,3,3,5,4,4
etc
In the example reference data above, you can see that the individual identifiers are listed in the first column, the known age at death appears in the next column, and the states of the N variables to be used to estimate age at death are in the next N columns (N=7 variables using the acetabular methods). In the case of the third specimen, for example, this individual has identity 'C-28' and known age at death of 41 years; it is in state 2 for variable 1 and state 3 for variable 2 (etc). Descriptions of specimens in the reference collection continue like this, one per row, until the end of the file.
- Identification of the variables: IDADE2 uses Bayesian
methods to estimate age of death of skeletal specimens based on a collection of reference variables.
These variables have to be the same in your reference file and in the observations of your test sample. However, they do not need to be the same as the variables in our reference data, if you are not using the Rissech et al. (2006) or SanMillán-Rissech method (San-Millán et al., 2016) of acetabular age estimation. In other words, if you choose IDADE2-Option 2, you do not need to use the acetabular methods, but can use your own reference collection and your method of choice.
- Variables of our reference collection:
Our reference collections use the 7 variables of the Rissech method (Rissech et al., 2006) and the revised versions of the variables published in the SanMillán-Rissech method (San-Millán et al., 2016). The original variables described by Rissech and collaborators (Rissech et al., 2006) are the following:
V1: Acetabular groove.
V2: Acetabular rim shape.
V3: Acetabular rim porosity.
V4: Apex activity.
V5: Activity on the outer edge of the acetabular fossa.
V6: Activity of the acetabular fossa.
V7: Porosities of the acetabular fossa.
The definitions of variables V1-V4 used in the SanMillán-Rissech method (San-Millán et al., 2016) are the same as in the original method with some small modifications. However, variables V5, V6 and V7 of SanMillán-Rissech method have different definitions than that of the original (see the Benchmark section).
V6: Texture and bone density in the cente of the acetabular fossa.
V7: Activity in the acetabular fossa.
For an accurate definition of all of these variables, see the original publications of these authors.
You can find their complete references in the Benchmark section. Please be certain that the acetabular reference data you select in IDADE2-Option1 corresponds with the acetabular aging method you used in your research.
- File descriptions of test specimens:
Test specimens have to be presented
in a file formatted identical to the file
with descriptions of specimens in the reference collection, except
that the age at death need not be known, in which case it is
entered as 0.
Output
The output consists of a table for each specimen listing the estimated age at death, various measures of
accuracy and the estimating probability distribution over age classes.
|
More details about the columns of the output of the server are reported below, along with an example of output data. For more information about the generation of output data, see the Usage page.
ID: Identification of the specimen.
AD: Known age of death.
ED: Estimated age of death.
Error: Measure of accuracy calculated as the difference between the estimated age and the known age.
CI: Confidence interval
Sequence of one or more consecutive age classes containing at least 95% of the distribution.
ACP: Age class probability.
Estimating probability distribution over age classes in per-thousand. Each age class corresponds to an age interval:
19: 19-23
24: 24-28
29: 29-33
34: 34-38
39: 39-43
44: 44-48
49: 49-53
54: 54-58
59: 59-63
64: 64-68
69: 69-73
74: 74-78
79: 79-83
84: 84-88
89: 89-93
94: 94-98
|
ID
|
AD |
ED |
ERROR |
CI |
19 |
24 |
29 |
34 |
39 |
44 |
49 |
54 |
59 |
64 |
69 |
74 |
79 |
84 |
89 |
94 |
|
1 |
74 |
81.6 |
7.6 |
69-93 |
0 |
0 |
0 |
0 |
0 |
0 |
5 |
0 |
0 |
7 |
121 |
284 |
66 |
338 |
179 |
0 |
4 |
29 |
32.9 |
3.9 |
29-53 |
2 |
0 |
892 |
0 |
0 |
32 |
74 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
5 |
75 |
77.6 |
2.6 |
69-88 |
0 |
0 |
0 |
0 |
0 |
0 |
5 |
0 |
0 |
0 |
51 |
696 |
90 |
157 |
0 |
0 |
7 |
83 |
81.0 |
2.0 |
79-83 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
1000 |
0 |
0 |
0 |
|
|
|
|
