I'm having trouble converting a Raster to an Array

Question

I'm currently having some trouble converting a raster to an array. Ultimately I would like to convert a raster from an integer to a float32 so that I can run a gdal_calc however, I could not change the type properly in order to do this.

SO, I would appreciate any help that someone could provide. Here is my script, very short.

import gdal
import numpy as np
import math
import osgeo
import os
import scipy


# Open Rasters 

hvRaster = gdal.Open("C:\\Users\\moses\\Desktop\\Calc_Test\\IMG-HV-ALOS2110871010-160611-HBQR1.5RUA.img")
vhRaster = gdal.Open("C:\\Users\\moses\\Desktop\\Calc_Test\\IMG-VH-ALOS2110871010-160611-HBQR1.5RUA.img")


# Get Raster Band

hvRasterBand = hvRaster.GetRasterBand(1)
vhRasterBand = vhRaster.GetRasterBand(1)

# Convert Raster to Array

hvArray = np.array(hvRaster.GetRasterBand(1).ReadAsArray())
vhArray = np.array(vhRaster.GetRasterBand(1).ReadAsArray())

print hvArray
print vhArray

Thank you in advance,

Moses

Answer 1

The method ReadAsArray() will create a numpy.ndarray with a dtype of the raster dataset. Your goal is to transform an integer dtype to a float32. The simplest way to do this is to use the astype() method for ndarray.

# Convert Raster to Array

hvArray = hvRaster.GetRasterBand(1).ReadAsArray().astype(np.float32)
vhArray = vhRaster.GetRasterBand(1).ReadAsArray().astype(np.float32)

Answer 2

Disclaimer: I'm the author of the following library.

lidario.Translator can help you to transform a raster into a point cloud (np.array):

import lidario as lio
    
# Create a Translator object which will take a tif file and return a np.array
translator = lio.Translator("geotiff", "np")
    
# Translate the tif file and get the np.array
point_cloud = translator.translate("/path/to/file.tif")

Lidario use GDAL under the hood.