Print descriptive statistics of a raster or numpy array.

stats.py

#!/usr/bin/env python3

from osgeo import gdal
import numpy as np
import pandas as pd

def stats(raster, band = 1, nodata = None):
	"""Print descriptive statistics of a raster or numpy array"""
	if (type(raster) != str) and (type(raster) != np.ndarray):
		print("Error: input must be a raster filepath or numpy array.")
		return
	elif (type(band) != int) or (band <= 0):
		print("Error: band selection must be a positive integer.")
		return
	elif type(raster) == str:
		file = gdal.Open(raster)
		img = np.array(file.GetRasterBand(band).ReadAsArray())
		if nodata is None: nodata = file.GetRasterBand(band).GetNoDataValue()
		file = None
	elif len(raster.shape) == 3:
		if raster.shape[0] < band:
			print("Error: there are only {} bands.".format(raster.shape[0]))
			return
		else:
			img = raster[band - 1]
	else:
		img = raster
	num_rows, num_cols = img.shape
	img_data = img[img != nodata]
	del img
	img_min = '{0:.2f}'.format(img_data.min())
	img_max = '{0:.2f}'.format(img_data.max())
	img_mean = '{0:.2f}'.format(img_data.mean())
	img_std = '{0:.2f}'.format(img_data.std())
	stats = {
		'Band': [band], 
		'Cols': [num_cols], 
		'Rows': [num_rows], 
		'Min': [img_min], 
		'Max': [img_max], 
		'Mean': [img_mean], 
		'Std': [img_std], 
		'NoData': [nodata]
	}
	df = pd.DataFrame.from_dict(stats)
	df.index = ['']
	print(df)
	return

Examples

To use a GeoTIFF on disk as the input:

>>> f = '~/example.tif'
>>> stats(f)
  Band   Cols   Rows   Min     Max   Mean    Std NoData
     1  40000  40000  0.00  100.00  51.14  49.21   None
>>>

To use a numpy array as the input:

>>> import numpy as np
>>> arr = np.array([[7,0,5,0],[4,0,3,9],[0,1,8,0]])
>>> print(arr)
[[7 0 5 0]
 [4 0 3 9]
 [0 1 8 0]]
>>> stats(arr)
  Band  Cols  Rows   Min   Max  Mean   Std NoData
     1     4     3  0.00  9.00  3.08  3.30   None
>>>