saturationSummary.py

# script to plot a histogram of some images
from astropy.io import fits
import matplotlib.pyplot as pl
from collections import defaultdict
from scipy import stats
import numpy as np
import glob as g
import pymysql, os

# 20160131
# Camera 801 - Mean: 59578        Median: 59552
# Camera 802 - Mean: 58597        Median: 58356
# Camera 805 - Mean: 61259        Median: 61263
# Camera 806 - Mean: 48729        Median: 48644
# Camera 809 - Mean: 65069        Median: 65045
# Camera 810 - Mean: 65037        Median: 64990
# Camera 811 - Mean: 64518        Median: 64486
# Camera 812 - Mean: 65169        Median: 65152

max_vals=defaultdict(list)
mini_survey_dir='/ngts/staging/archive/minisurvey'

db=pymysql.connect(host='ds',db='ngts_ops')
qry="SELECT image_id,camera_id FROM mini_survey WHERE astrometry>0 AND done=1"
rc=1
image_id,camera_id=[],[]
with db.cursor() as cur:
	cur.execute(qry)
	for row in cur:
		image_id.append(row[0])
		camera_id.append(int(row[1]))
db.close()

for i in range(0,len(image_id)):
	with fits.open('%s/IMAGE%s.fits' % (mini_survey_dir,image_id[i])) as h:
		d=np.ravel(h[0].data)
		max_vals[camera_id[i]].append(max(d))
		img_name='%s/IMAGE%s.hist.png' % (mini_survey_dir,image_id[i])
		if os.path.exists(img_name)==False:
			fig,ax=pl.subplots(1,figsize=(10,10))
			ax.hist(d,range(1,67000)[::100])
			ax.set_yscale('log', nonposy='clip')
			fig.savefig(img_name,dpi=200)
			pl.close()
		else:
			print('%s exists, skipping...' % (img_name))
		print rc
		rc+=1

# now make a histogram of the max values per camera
c=0
cameras=np.arange(801,813,1)
figm = pl.figure(rc+1,figsize=(15,15))
figm.clf()
for i in cameras:
	ax = figm.add_subplot(4, 3, c+1)	
	ax.set_title(i)
	if i in max_vals:
		ax.hist(max_vals[i],range(40000,67000)[::100])
		ax.set_yscale('log', nonposy='clip')
	c+=1
figm.savefig('%s/saturationSummary.png' % (mini_survey_dir))
pl.close()

# loop over the values and print the mean and median
for i in max_vals:
	print "Camera %d - Mean: %d\tMedian: %d" % (i,np.average(max_vals[i]),np.median(max_vals[i]))