graphing.py

import matplotlib.pyplot as plt
from f_bidr import *
from f_bidr_data import get_orbit_file_path as orbit
from images import *
import itertools

import numpy as np

# This file does nothing in particular. This is to assist graphical
# exploration of the FBIDR records, and do small work. In contrast to
# images.py which is for producing pictures of full orbits and mosaics
# of multiple orbits, which may need as much memory as it can have, so
# things are cut out of images.py.
# 
# It's best to run this interactively using python -i graphing.py
# and just play around.

def get(records, *names):
    """For extracting per-record information from a list of orbits.
    Returns a list of data, or list of lists of data."""
    outputs = []
    for name in names:
        if callable(name):
            outputs.append([name(r) for r in records])
        else:
            outputs.append([r[name] for r in records])
    if len(names) == 1:
        return outputs[0]
    else:
        return outputs

# record_iter is an iterable of lists of records. For help with
# comparing different orbits.
def gets(record_iter, name):
    """Name here is limited to one name. Gets one field from multiple
    collections of records."""
    return [get(lst, name) for lst in record_iter]

def graph(records, *names, **axargs):
    stuff = get(records, *names)
    if len(names) == 1:
        plt.scatter(range(len(stuff)), stuff)
        ax = plt.gca()
        ax.set(xlabel='record #', ylabel=names[0])
    else:
        plt.scatter(stuff[0], stuff[1])
        ax = plt.gca()
        ax.set(xlabel=names[0], ylabel=names[1])
    if axargs:
        ax.set(**axargs)
    plt.show()

def find_first(lst, test):
    """Helpful for finding records within a particular lat/lon range."""
    for i in range(len(lst)):
        if test(lst[i]): 
            return i
def read_orbit(*args, num_records=None):
    return read_logical_records(orbit(*args), num_records)

def measure_overlap(records):
    line_offsets = np.array(get(records, 'reference_offset_lines'))
    heights = np.array(get(records, 'line_count'))
    # l0 - l1, l1 - l2, l2 - l3
    differences = line_offsets[:-1] - line_offsets[1:]
    return differences - heights[:-1]