Raster Correlation

The library provides an easy way correlate raster of the same size. It returns a json containing statistical outputs and frequencies information to be directly used with Highcharts JS or Matplotlib chart libraries.

Installation

Dependencies

The library has different dependencies (see also requirements.txt) click, watchdog, flask, flask-cors, numpy, scipy, pysal, brewer2mpl, rasterio, GeobricksCommon.

On Ubuntu

sudo add-apt-repository ppa:ubuntugis/ppa
sudo apt-get update
sudo apt-get install python-numpy libgdal1h gdal-bin libgdal-dev

In case of compiling errors for numpy

sudo apt-get install libblas3gf libc6 libgcc1 libgfortran3 liblapack3gf libstdc++6 build-essential gfortran python-all-dev libatlas-base-dev python-dev

In case of compiling errors for scipy

sudo apt-get install libblas-dev liblapack-dev

Installation

The library is distributed through PyPi and can be installed by typing the following commands in the console:

pip -r https://raw.githubusercontent.com/geobricks/geobricks_raster_correlation/master/requirements.txt

pip install GeobricksRasterCorrelation

N.B. Due to a well known PyPi issue it's not possible to install scipy and pysal through setup.py or requirements.txt

In order to install pysal run the following command

pip install pysal

Examples

Library usage

from geobricks_raster_correlation.core.raster_correlation_core import get_correlation

raster_path1 = "path_to_raster1.tif"
raster_path2 = "path_to_raster2.tif"
# Number of bins to be applied to the scatter chart
bins = 300
corr = get_correlation(raster_path1, raster_path2, bins)
print corr

Example with matplotlib

This example generate a correlation chart with matplotlib

from geobricks_raster_correlation.core.raster_correlation_core import get_correlation
from matplotlib import pyplot as plt
from matplotlib.pylab import polyfit, polyval

# input to your raster files
raster_path1 = "path_to_raster1.tif"
raster_path2 = "path_to_raster2.tif"

# Number of sampling bins
bins = 150

corr = get_correlation(raster_path1, raster_path2, bins)
x = []
y = []
colors = []
# print corr['series']
for serie in corr['series']:
    colors.append(serie['color'])
    for data in serie['data']:
        x.append(data[0])
        y.append(data[1])

# Adding regression line
(m, b) = polyfit(x, y, 1)
yp = polyval([m, b], x)
plt.plot(x, yp)

# plotting scatter
plt.scatter(x, y, c=colors)
plt.show()

The returned json:

• corr['stats'] contains the statistics: slope, p_value, std_err, intercept, r_value
• corr['series'] contains the output series that can be used directly as an Highcharts input or with Matplotlib.