Python interface to SLEUTH¶

An object oriented wrapper for SLEUTH, an urban growth model.

This module defines the Location class, which is a work unit for SLEUTH.

How to use this module¶

Import it:
```
import sleuth_automation as sa
```

Configure it at least with a path to SLEUTH binaries:

sa.configure(sleuth_path='/path/to/sleuth',
             use_mpi=True, mpi_cores=32)

Create an instance:

l = sa.Location('my_location',
                '/path/to/my_location')

Calibrate model:

l.calibrate_coarse()
l.calibrate_fine()
l.calibrate_final()

or calibrate with a single call:

l.sleuth_calibrate()

Predict growth:
```
l.sleuth_predict(2017, 2060)
```
Format prediction as TIFF:
```
l.gif2tif(2017, 2060)
```

The Location class¶

class sleuth_automation.Location(location, input_path)[source]¶

This class represents the main work unit of the SLEUTH model. Within a location urban growth is simulated. A location should have a name which is just a string and is used for naming output files, and a directory containing GIF files as specified by the SLEUTH documentation.

calibrate_coarse(monte_carlo_iterations=50)[source]¶

Run SLEUTH coarse calibration.

Parameters:

monte_carlo_iterations: iterations for the coarse calibration step.

calibrate_final(monte_carlo_iterations=50)[source]¶

Run SLEUTH final calibration.

Parameters:

monte_carlo_iterations: iterations for the final calibration step.

calibrate_fine(monte_carlo_iterations=50)[source]¶

Run SLEUTH fine calibration.

Parameters:

monte_carlo_iterations: iterations for the fine calibration step.

create_scenario_file(params, monte_carlo_iterations)[source]¶

Merges extra parameters to global config and renders them into a SLEUTH scenario file.

Parameters:

params: extra parameters to merge.
monte_carlo_iterations: more iterations take a longer time

to run but add precision.

gif2tif(start, end)[source]¶

Convert prediction output from GIF to TIFF.

Will use extent.json and gdal_translate to recover geospatial information lost to GIF format.

Parameters:

start: beginning of prediction range.
end: ending of prediction range, inclusive.

Start and end parameters are necessary to figure out paths.

sleuth_calibrate()[source]¶: Run coarse, fine and final calibration steps in one call.

sleuth_predict(end, diff=None, brd=None, sprd=None, slp=None, rg=None, monte_carlo_iterations=50)[source]¶

Run SLEUTH prediction step for range enclosed in start, end.

Parameters:

end: end of temporal range for prediction, beggining is infered from last input
diff: TODO: see model documentation
brd: TODO: see model documentation
sprd: spread
slp: slope
rg: TODO: see model documentation
monte_carlo_iterations: iterations for the prediction step

ControlStats module¶

An object representation of a SLEUTH control_stats.log file.

control_stats.log files are the output of calibration steps. They contain the ranges for next calibration or prediction steps. This helper module extracts those ranges for all SLEUTH parameters:

diff

brd

sprd

slp

rg

This is helper module will extract parameter ranges into a dictionary that can be rendered into a jinja2 template for the creation of SLEUTH scenario files.

How to use this module¶

Import it: from controlstats import ControlStats.

Create an instance:

cs = ControlStats('/path/to/my_location/output/coarse/control_stats.log'),
                  step=5)

parameter ranges are extracted into the params attribute: cs.params

cs.params is passed to Location.create_scenario_file to render scenario files.

ConstrolStats class¶

class sleuth_automation.controlstats.ControlStats(path, step)[source]¶

This class grabs values from a control_stats.log file and calculates ranges for the parameters for next calibration stage.

It will find ranges for:

diff
brd
sprd
slp
rg

Including values for start, end and step.

All fetched parameters will be stored into the params property.