MonitoringOutput

Output of simulations

Output from OpenMalaria simulations is currently collected using two methods. The monitoring page describes how these methods work and are configured, while the monitoring measures page lists what data these methods are able to output.

Surveys

Output from the surveys system is written at the end of the simulation to the output.txt (or when compiled in BOINC mode output.txt.gz) file. The contents are formatted as in a database, with tab-separated columns and one record per line. Each line has four columns with the following meaning:

survey number	third dimension	measure	value

The survey number starts from one and corresponds to the survey time point. (Exception: measure 21 has one record from the end of the simulation and does not use the survey number or third dimension columns.)

Output can be associated with several different measures; the code under the label "id" in the first column of the survey measures table appears in the third column of output.

The "third dimension" (in the second column for historical reasons) specifies another dimension of the output. For many measures it identifies the human age group, for a few measures it is unused, and for some it holds a mosquito species or a drug identifier.

Example:

1       15      2       4

Measure 2 is the "expected number of infected hosts", so this record specifies that during the first survey, the expected number of infected hosts within the fifteenth age group was 4.

Continuous monitoring

Some properties of the simulation can also be monitored continuously. The continuous monitoring system writes output to the ctsout.txt file in real-time (while the simulation is still running). This file can be monitored while a simulation is running from LiveGraph and read from most spread-sheet programs.

This file starts with a line specifying the column separator (a tab) to make the files readable by LiveGraph. The next line is a header, giving a title for each column (note: spaces may occur within column names, whereas tab characters separate columns). Following this, each line corresponds to a time point.

Available outputs are listed here; note that some of these output a single value while others output multiple values (using titles of the format "measure (subcategory)").