# -*- coding: utf-8 -*-

"""

Created on Mon Jun 13 15:21:46 2016

@author: throop

"""

import hbt

import os

import pickle

import astropy

import spiceypy as sp

import matplotlib.pyplot as plt

from matplotlib.backends.backend_pdf import PdfPages

import numpy as np

##########

# NH_Jring_invengory

# Take an inventory of all of the NH Jupiter rings images.

# Make a PDF file showing all of the images, along with a bit of info.

# Also, make a text file with tabular data.

#

# HBT 14-Jun-2016

# HBT 2-May-2017 Updated to Python3, and for new directory structure.

#

##########

# Possible additions: Phase angle [DONE]

# Distance from Jupiter.

# Time before/after C/A

filename_save = 'nh_jring_read_params_5711.pkl' # Filename to save parameters in

#filename_save = 'nh_jring_read_params_100.pkl' # Filename to save parameters in

file_tm = "/Users/throop/gv/dev/gv_kernels_new_horizons.txt" # SPICE metakernel

dir_images = '/Users/throop/data/NH_Jring/data/jupiter/level2/lor/all/'

dir_out = '/Users/throop/Data/NH_Jring/out/'

file_out_pdf = 'nh_jring_inventory.pdf' # PDF output file which is created

file_out_txt = file_out_pdf.replace('pdf', 'txt')

sp.furnsh(file_tm) # Commented out for testing while SPICE was recopying kernel pool.

# Check if there is a pickle save file found. If it is there, go ahead and read it.

if os.path.exists(filename_save):

print("Loading file: " + filename_save)

# self.load(verbose=False) # This will load self.t

lun = open(filename_save, 'rb')

t = pickle.load(lun)

# self.t = t # All of the variables we need are in the 't' table.

lun.close()

# t = self.t

# Find and process all of the FITS header info for all of the image files

# 't' is our main data table, and has all the info for all the files.

else:

# Search for all files. Ideally I would search for all files that don't have 'opnav',

# but the way the filter is set up, it is easier to search for all that do. Result is the same.

t = hbt.get_fits_info_from_files_lorri(dir_images, pattern='opnav')

num_files = np.size(t)

print('Read ' + repr(num_files) + ' files.')

# XXX Now shorten the list arbitrarily, just for testing purposes

#t = t[100:112] # XXX shorten it! 100 does not get to 4x4 territory

# Get a list of unique observation descriptions (i.e., 'groups')

groups = astropy.table.unique(t, keys=(['Desc']))['Desc']

num_groups = np.size(groups)

# Set initial location to first file in first group

index_group = 0

index_image = 0

num_rows = 3

num_cols = 4

# Set the starting row & column position

row = 1

col = 1

i = 1 # Current image #, starting at 1

i_on_page = 1 # Current image #, on this page

i_page = 1 # Current page number, within this group

i_file = 0 # Index of this file, within the group

i_group = 0 # Index of this group, within the list of groups

lines_out = [''] # This array is the output text file. It will be written monolithically at the end.

# Less chance of leaving open.

plt.rc('image', cmap='Greys_r') # Use a non-inverted colormap: White = stars and ring, etc.

# Start the PDF

pp = PdfPages(dir_out + file_out_pdf)

fs = 5.5 # Font size for the PDF. 2.7 is very tiny. 7 is good but text is a bit too large per cell.

# Now loop over every image

for i_group,group in enumerate(groups):

mask = t['Desc'] == group

files = t[mask]

dt = 0

t0 = 0

header = " ----- " + group + " ----- "

print

print(header)

lines_out.append('')

lines_out.append(header)

# Starting a new group, we start in UL corner on new page

row = 1

col = 1

i_on_page = 1

i_page = 1

# print("Generating output...")

for i_file,file in enumerate(files): # Loop over files in this particular group

if (t0 ==0): # Calculate time since previous image

dt_s = 0

else:

dt_s = float(file['ET']) - t0

m, s = divmod(dt_s, 60)

h, m = divmod(m, 60)

dt_str = "{:3d}h {:2d}m {:2d}s".format(int(h), int(m), int(s))

if (dt_s == 0): dt_str = '--'

dt_str = dt_str.replace(' 0h', '').replace(' 0m', '')

t0 = float(file['ET'])

# Create a super-short version of the filename (cut out the ApID)

file_trunc = file['Shortname'].replace('lor_', '').replace('_0x630_sci', '').\

replace('_0x633_sci', '').replace('_opnav', '').replace('.fit', '')

utc_trunc = sp.et2utc(sp.utc2et(file['UTC']),'C', 0)

# Print a line of the table, to the screen and the file

line = "{:>3}/{:<3}: {:>3}, {:>3}, {}, {}, {}, {:6.3f},{:>12s}, {:.1f} deg, {:<9}".format(int(i),

int(num_files), int(i_group),

int(i_file), file_trunc, file['Format'], utc_trunc,

file['Exptime'], (dt_str), file['Phase']*hbt.r2d, file['Target'])

print(line)

lines_out.append(line)

arr = hbt.read_lorri(file['Filename'], bg_method = 'Polynomial', bg_argument = 4, frac_clip = 1)

arr = hbt.remove_brightest(arr, 0.99)

arr = -hbt.remove_brightest(-arr, 0.99)

# Plot the image to the PDF

p = plt.subplot(num_rows, num_cols,i_on_page) # Image 1

plt.imshow(arr, interpolation='none')

# plt.tight_layout() # Reduces whitespace, but in this case pushes captions off the edge of page!

a = plt.gca() # Get the axes

a.get_xaxis().set_visible(False) # Do not plot the axes

a.get_yaxis().set_visible(False)

n1 = int(file['N1']) # NAXIS1: This will be either 1024 [1x1], or 255 [4x4]

scalefac = n1 / 1024. # Generate a scaling factor. This is 1.0 for a normal image,

# or 0.25 for a 4x4. Use it for placing text() properly.

# Generate the text to put next to each image on the PDF

label1 = "{}/{}: {}, {}, {} s; {}".format(int(i), int(num_files), file_trunc, file['Format'],

file['Exptime'], dt_str)

label2 = r'{}, {:.1f}$^\circ$, Group {}, File {}'.format(utc_trunc, file['Phase']*hbt.r2d,

int(i_group), int(i_file))

plt.text(0, -90*scalefac, label1, fontsize = fs)

plt.text(0, -30*scalefac, label2, fontsize = fs)

# Generate the text to be on the header of each page

if (i_on_page == 1):

plt.text(-250*scalefac, 200*scalefac, group + ', group ' + repr(i_group) + ' ' +

" page " + repr(i_page),

fontsize=fs*2, rotation=90)

# print "Just plotted string: " + str

# Now update the row/column for the next image

i += 1

i_on_page += 1

col = np.mod(i_on_page-1, num_cols)+1 # Calc new column number. The +1 is since they go 1, 2, 3 not 0, 1, 2

if (col == 1):

row += 1

if (row == num_rows+1): # If we have just filled up the current page, and are starting a new one

# print "NEW PAGE! Page full."

fig = plt.gcf()

pp.savefig(fig) # Close the page, and start a new one, with the same group

fig.clf()

i_on_page = 1 # Start the new images at UL corner of new page

i_page += 1

row = 1

col = 1

# print

if (i_on_page != 1):

fig = plt.gcf() # If we have finished the loop for the current group, start a new page.

# (NB this will double-paginate for N=12 files.)

pp.savefig(fig)

fig.clf()

# print "NEW PAGE! Done with group."

# plt.imshow(arr)

# plt.show()

#nh_jring_inventory()

# http://stackoverflow.com/questions/2252726/how-to-create-pdf-files-in-python

#pp.savefig(plt.gcf()) # This generates page 1

print("Writing pdf...")

pp.close()

print("Wrote: " + dir_out + file_out_pdf)

print("Writing txt...")

np.savetxt(dir_out + file_out_txt, np.array(lines_out), fmt='%s')

print("Wrote: " + dir_out + file_out_txt)

print('Processed ' + repr(np.size(t)) + ' files.')

# Now make some plots of various quantities vs. time

# For ring itself (excluding the Io 'monitoring' obs), the total is N = 517 obs.

# Including the erroneous 'monitoring' ones, N = 571.

# And then thre is also LORRI and anything at random. And Himalia ring of Cheng et al.

jca_utc = '2007 FEB 28 05:41:48' # Looked up from GV

jca_et = sp.utc2et(jca_utc)

groups = astropy.table.unique(t, keys=(['Desc']))['Desc']

for i_group,group in enumerate(groups):

plt.subplots(1, 3, figsize=(13,4))

mask = t['Desc'] == group

t_days = (t['ET'][mask] - jca_et) / 86400

phase = t['Phase'][mask] * hbt.r2d

dist = t['dxyz'][mask] / 71492 # Distance from Jup, km

res = 0.3 / 1024 * hbt.d2r * t['dxyz'][mask] # Spatial resolution, km

plt.subplot(1,3,1)

plt.plot(t_days, phase, marker = 'o', linestyle='none', label=group)

plt.xlabel('Days from Jupiter C/A')

plt.ylabel('Phase Angle')

plt.xlim((-5,4))

plt.ylim((00,180))

plt.subplot(1,3,2)

plt.plot(t_days, dist, marker = 'o', linestyle='none')

plt.title(group + ', n=' + repr(np.sum(mask)))

plt.xlim((-5,4))

plt.ylabel('Distance [RJ]')

plt.xlabel('Days from Jupiter C/A')

plt.ylim((20, 110))

plt.subplot(1,3,3)

plt.plot(t_days, res, marker = 'o', linestyle='none')

plt.xlim((-5,4))

plt.ylabel('Resolution [km]')

plt.xlabel('Days from Jupiter C/A')

plt.ylim((10, 110))

plt.show()

plt.plot((t['ET'] - jca_et) / 86400, t['Exptime'],linestyle='none', marker = '+')

plt.xlabel('Days from Jupiter C/A')

plt.ylabel('Exptime [s]')

# arr = hbt.read_lorri(dir_images + '/lor_0035122328_0x633_sci_1.fit')

# arr = hbt.read_lorri(dir_images + '/lor_0035282790_0x633_sci_1.fit')

stop