Metrics

The following metrics are currently available on demcompare:

mean
max
min
std (Standard Deviation)
rmse (Root Mean Squared Error)
median
nmad (Normalized Median Absolute Deviation) = $1.486*median(\lvert data - median(data)\rvert)$
sum
squared_sum
percentil_90

Name	Type	Parameters	Type	Default value
`'cdf'`Cumulative Density Function	vector	bin_step	float	`0.1`
		output_csv_path	string	`None`
		output_plot_path	string	`None`
`'pdf'`Probability Density Function	vector	bin_step	float	`0.2`
		width	float	`0.7`
		filter_p98	boolean	`false`
		output_csv_path	string	`None`
		output_plot_path	string	`None`
`'ratio_above_threshold'`	vector	elevation_threshold	List[float, int]	$[0.5, 1, 3]$
		original_unit	string	`"m"`
		output_csv_path	string	`None`
`'slope-orientation-histogram'`	vector	output_plot_path	string	`None`

Name	Type	Parameters	Type	Default value
`'hillshade'`Hill shade	matrix	azimuth	float	`0.9`
		angle_altitude	float	`45`
		cmap	str	`Greys_r`
		cmap_nodata	str	`royalblue`
		colorbar_title	str	`Hill shade`
		fig_title	str	`DEM hill shade`
		plot_path	str	`None`
`'svf'`SkyViewFactor	matrix	filter_intensity	float	`315`
		replication	bool	`True`
		quantiles	List[float]	$[0.09, 0.91]$
		cmap	str	`Greys_r`
		cmap_nodata	str	`royalblue`
		colorbar_title	str	`Sky view factor`
		fig_title	str	`DEM sky view factor`
		plot_path	str	`None`

Note

The metrics are always computed on valid pixels. Valid pixels are those whose value is different than NaN and the nodata value (-32768 by default if not specified in the input configuration or in the input DEM).

Note

Apart from only considering the valid pixels, the user may also specify the remove_outliers option in the input configuration. This option will also filter all DEM pixels outside (mu + 3 sigma) and (mu - 3 sigma), being mu the mean and sigma the standard deviation of all valid pixels in the DEM.

Note

'ratio_above_threshold' and 'slope-orientation-histogram' are not computed by default. They must be indicated in the configuration file in order to be used. An example on how to include them in the configuration is shown below.

Note

'slope-orientation-histogram' should have the 'output_plot_path' parameter specified, otherwise the plot will not be saved. An example on how to include it in the configuration is shown below.

Note

More informations about the hillshade and the sky-view factor can be found in Hillshade and sky-view factor representations.

{
    "output_dir": "./test_output/",
    "input_ref": {
        "path": "./Gironde.tif",
        "nodata": -9999.0
    },
    "input_sec": {
        "path": "./FinalWaveBathymetry_T30TXR_20200622T105631_D_MSL_invert.TIF",
        "nodata": -32768
    }
    "statistics": {
        "alti-diff": {
            "remove_outliers": "True",
            "metrics": ["mean", {"ratio_above_threshold": {"elevation_threshold": [1, 2, 3]}}]
        },
        "ref": {
            "remove_outliers": "True",
            "metrics": [
                {
                    "slope-orientation-histogram": {
                        "output_plot_path": "path_to_plot"
                    }
                }
            ]
        }
    }
}

Note

'slope-orientation-histogram'’s plots are not saved in the report.

Warning

The combination of DEM processing methods and metrics may not be meaningful!