1_3_Raster_images_as_numpy_arrays.ipynb

{"cells":[{"cell_type":"markdown","source":["A raster graphic represents a two-dimensional image as a rectangular matrix or grid of square pixels. A raster is technically characterized by the width and height of the image in pixels and by the number of bits per pixel.\n","\n","Geographic phenomena are commonly represented in a raster format in GIS. The raster grid is georeferenced, so that each pixel (commonly called a cell in GIS because the \"picture\" part of \"pixel\" is not relevant) represents a square region of geographic space. The value of each cell then represents some measurable (qualitative or quantitative) property of that region, typically conceptualized as a field. Examples of fields commonly represented in raster include: temperature, population density, soil moisture, land cover, surface elevation, etc.\n","\n","Read about satellite images interpretation: https://earthobservatory.nasa.gov/features/ColorImage"],"metadata":{"id":"QnxClC0ZpYtJ"}},{"cell_type":"code","source":["from google.colab import drive\n","drive.mount('/content/drive')"],"metadata":{"id":"rITDDriP5zY2"},"execution_count":null,"outputs":[]},{"cell_type":"code","execution_count":null,"metadata":{"id":"tip5mjOMg1Fg"},"outputs":[],"source":["## numpy arrays are faster than lists!\n","## but can be harder for beginners to comprehend\n","## I see rasters as numpy arrays\n","## data cubes will display in 4 dimensional numpy arrays!\n","\n","import numpy as np\n","\n","# 2D arrays [ROWS, COLUMNS]"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"98SVr9t3g1Fj"},"outputs":[],"source":["np.random.seed(100)\n","\n","directory = '/content/drive/MyDrive/GIS_tutorials/2_python_main_gis_functions/data_package_beg/'\n","fileName = directory + \"out1.tif\"\n"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"s8yklW26g1Fk"},"outputs":[],"source":["## create a random array\n","array1d = np.random.randint(25, size=10)\n","\n","print(array1d)\n","\n","## looks like a list"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"_ry7DLPWg1Fl"},"outputs":[],"source":["print(array1d.shape)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"o3fz2aDxg1Fl"},"outputs":[],"source":["print(array1d[0])\n","\n","## behaves like a list"]},{"cell_type":"markdown","metadata":{"id":"BtyuBHhRg1Fm"},"source":["Starting to look like an image?"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"r9Y3T9JEg1Fn"},"outputs":[],"source":["array2d = np.random.randint(25, size=(10,12))\n","print(array2d)\n","print(array2d.shape)\n","print(array2d[0][0])"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"uzlwFfZdg1Fo"},"outputs":[],"source":["## datatype can also be useful\n","\n","print(array2d.dtype)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"INC6ltxpg1Fp"},"outputs":[],"source":["## slicing!\n","\n","array2d[:2,:5]    ## two rows, five columns - use the above to visualise"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"EkFvyxkUg1Fp"},"outputs":[],"source":["array2d[:,0]    ## everything in the first column"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"ZvwVXHqUg1Fq"},"outputs":[],"source":["array2d[0,:]    ## everything in the first row"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"1mMmAwuhg1Fq"},"outputs":[],"source":["## lets print the array again...\n","print(array2d)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"-gokv_-4g1Fr"},"outputs":[],"source":["## ...and change some values\n","array2d[array2d > 10 ] = 1"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"q0i94Nwfg1Fr"},"outputs":[],"source":["print(array2d)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"AWahKRRvg1Fs"},"outputs":[],"source":["## or\n","a = np.random.randint(25, size=(12,12))\n","b = np.where(a>10,1,a)\n","\n","print(a)\n","print (\"\")\n","print(b)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"E6bvd6jmg1Fs"},"outputs":[],"source":["%matplotlib inline\n","### Graphing is a very useful tool in python and matplotlib is the workhorse here\n","### Graphing is a very visual way of quickly seeing what is going on with your data\n","\n","import matplotlib.pyplot as plt\n","\n","plt.imshow(a)\n","plt.show()"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"Qwfr0DD7g1Fs"},"outputs":[],"source":["plt.imshow(b)\n","plt.show()"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"6JnrAWBMg1Ft"},"outputs":[],"source":["### we can also do some pretty cool stuff fast\n","## This is known as broadcasting\n","\n","c = a + b\n","plt.imshow(c)\n","plt.show()"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"pAp1MnX2g1Ft"},"outputs":[],"source":["print(a)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"-JKxLoVhg1Ft"},"outputs":[],"source":["## reshape the array\n","d = a.flatten()\n","print(d.shape)\n","print(d)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"btdTsov8g1Ft"},"outputs":[],"source":["## reshape it back\n","e = d.reshape(12,12)\n","print(e.shape)\n","print(e)"]},{"cell_type":"markdown","source":["Challenge! Can you create a masked image of 0,1 on a grid 12,12, where <10 set to 0 where >=10 set 1"],"metadata":{"id":"A9hFMmhDq93O"}},{"cell_type":"code","execution_count":null,"metadata":{"id":"s0u2SIPJg1Fu"},"outputs":[],"source":["### or... masking\n","k = np.random.randint(25, size =(12,12))\n","\n","l = np.where(k<10,0,k)\n","l = np.where(k>=10,1,l)\n","\n","print(l)\n","\n","plt.imshow(l)\n","plt.show()\n"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"vMW8mW-Og1Fu"},"outputs":[],"source":["##challenges!\n","# create a numpy array of a 3X3 matrix with values 0-8\n","\n","test = np.arange(9)\n","print(test)\n","print()\n","\n","array_3 = np.arange(9).reshape(3,3)\n","print(array_3)\n","\n"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"DcVEV16og1Fu"},"outputs":[],"source":["## break it down if you are not sure\n"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"F_eD7sUkg1Fv"},"outputs":[],"source":["### numpy arrays are very powerful and fast. This only touches on them. \n","##rasters are 2d or 3d arrays. We can perform multiple functions on a raster with numpy and its fast"]},{"cell_type":"markdown","metadata":{"id":"irnsFfs2g1Fv"},"source":["# Moving onto rasters"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"7Mr9k9M_g1Fv"},"outputs":[],"source":["%matplotlib inline\n","import gdal     # this lib will be discussed in next notebook\n","import matplotlib.pyplot as plt\n","\n","image_path = fileName\n","\n","raster_ds = gdal.Open(image_path, gdal.GA_ReadOnly)\n","image_gdal = raster_ds.GetRasterBand(1).ReadAsArray() ## only a single band though\n","image_gdal2 = raster_ds.GetRasterBand(2).ReadAsArray() ## only a single band though\n","image_gdal3 = raster_ds.GetRasterBand(3).ReadAsArray() ## only a single band though\n","\n","\n","print(image_gdal.shape) \n","print(\"It's a numpy array!\")\n","\n","\n","plt.imshow(image_gdal, cmap = \"Reds\")\n","plt.show()\n","\n","plt.imshow(image_gdal2, cmap = \"Greens\")\n","plt.show()\n","\n","plt.imshow(image_gdal3, cmap = \"Blues\")\n","plt.show()"]},{"cell_type":"markdown","metadata":{"id":"FChs-3mqg1Fv"},"source":["This is not a good idea!"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"fzoLE4Epg1Fv"},"outputs":[],"source":["## what about iterating over all the pixels and printing them?\n","print(image_gdal)\n","\n","for pixel in image_gdal:\n","    print(pixel)"]},{"cell_type":"markdown","metadata":{"id":"r908e3QBg1Fw"},"source":["Normalizing"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"XTkgrF1Ag1Fw"},"outputs":[],"source":["Zmax, Zmin = image_gdal.max(), image_gdal.min()\n","Z = (image_gdal - Zmin)/(Zmax - Zmin)\n","print(Z)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"BmYRd5pHg1Fw"},"outputs":[],"source":["plt.imshow(Z)\n","plt.show()"]},{"cell_type":"markdown","metadata":{"id":"tgX0Ljprg1Fw"},"source":["Subtraction"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"xObxRM7Yg1Fw"},"outputs":[],"source":["## Subtract the mean of each row of a matrix \n","Y = Z - Z.mean(axis=1, keepdims=True)\n","plt.imshow(Y)\n","plt.show()"]}],"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.7.9"},"colab":{"name":"1_3_Raster_images_as_numpy_arrays.ipynb","provenance":[],"collapsed_sections":[]}},"nbformat":4,"nbformat_minor":0}