 |
This page briefly explains the format that must be
followed when creating a new population file to use with the Human
Measurements Editor product. Any file extension will do, but the
extension .sws (for Safework
Statistic) is
usually reserved for these types of files. Once the file is created, it
possible to load it into V5 using the steps described in Basic Task
User-defined Population Databases.
A population file is organized into sections of data; each section must begin with a keyword and end with a keyword. The ending keyword of one section is the beginning keyword of the next section, except for the last section where the ending keyword must be 'END'. All blank lines are discarded, and all lines beginning with the '!' (exclamation point) character are considered as comments, and are also discarded. A population file may contain at most four sections, with the following keywords:
All sections are optional; the MEAN_STDEV sections must appear before the CORR sections. Furthermore, a given keyword may not appear twice in the same population file. In the MEAN_STDEV sections, the user may provide values (mean and standard deviation) of every measurement reflecting the population of the study. There must be one line per entry, and each entry must describe, at most, one variable in the following fashion: <variable> <mean> <stddev> where <variable> is the variable reference number, <mean> is the mean value of the variable and <stddev> is the standard deviation value defining that variable. A population file may be as simple as the following:
In the population file above, the mean value of the stature (variable us100) of a male manikin is defined as 177 centimeters (70 inches), with a standard deviation of 6.0. Likewise, the mean stature value of a female manikin is 164 centimeters (64.5 inches). In the CORR sections, the user may provide correlation values between any pair of variables. The correlation between two variables can be defined as a real number in the range [-1.0 ; 1.0], expressing the relative dependency between the two variables. The higher the correlation absolute value, the more dependant the variables are on each other. When defining correlations, there must be one line per entry, and each entry must describe one correlation between one pair of variables, in the following fashion: <variable1> <variable2> <correlation> where <variable1> is the reference number of the first variable, <variable2> is the reference number of the second variable, and <correlation> is the correlation value linking the two variables together. Please note that <variable1> must be different from <variable2>, because by definition, the correlation between a variable and itself is always 1.0. Furthermore, the reference number of <variable1> must be lower than the reference number of <variable2> If the correlation value given is not within the range [-1.0; 1.0], then an error is raised. The following file gives an example of correlation specifications:
Please note that all length values appearing in a population file must be provided in centimeters, and all weight values must be provided in kilograms. Also note that the keywords in a population file are case sensitive. As such, the keyword mean_stdev f would be considered a syntax error. |
||||||||||||||||||||||||||||||||||||||||||||||||
 |
An example of a population file, my_population.sws, is provided in the samples directory. Also appearing in the samples directory are the five default populations available at manikin creation: |
||||||||||||||||||||||||||||||||||||||||||||||||
|
|
Please refer to Anthropometric Variables for more information. This table contains all information about each variable used in Human Measurements Editor including the reference number, the acronym, the full name, and the definition of each variable. | ||||||||||||||||||||||||||||||||||||||||||||||||
