azely#

Release Python Downloads DOI Tests

Computation and plotting of azimuth and elevation for astronomical objects

TL;DR#

Azely (pronounced as “as-elie”) is a Python package for computation and plotting of horizontal coordinates (azimuth and elevation; az/el, hereafter) of astronomical objects at given location and time. While computation and plotting are realized by Astropy and Matplotlib, what azely provides is high-level API to use them easily. In fact azely offers one-liner to compute and plot, for example, one-day elevation of the Sun in Tokyo:

>>> azely.compute('Sun', 'Tokyo').el.plot(ylim=(0, 90))

one-liner.svg

Features#

  • High-level API: azely provides a simple yet powerful compute() function. Users can complete most of operation with it (e.g., information acquisition and computation).

  • Handy output: azely’s output (from compute()) is pandas DataFrame, a de facto standard data structure of Python. Users can convert it to other formats like CSV and plot it by Matplotlib using builtin methods.

  • Web information acquisition: azely can automatically acquire object and location information (i.e., longitude and latitude) from online services (e.g., catalogues or maps). Obtained information is cached in a local TOML file for an offline use.

  • User-defined information: azely also offers to use user-defined object and location information written in a TOML file.

Requirements#

  • Python: 3.7, 3.8, 3.9, and 3.10 (tested by author)

  • Dependencies: See pyproject.toml

Installation#

$ pip install azely

Basic usage#

This section describes basic az/el computation using compute() function.

Compute function#

Azely’s compute() function receives the following parameters and returns pandas DataFrame (df):

>>> import azely
>>> df = azely.compute(object, site, time, view, **options)

This means that azely will compute az/el of object observed from site at (on) time in view. For example, the following code will compute az/el of Sun observed from ALMA AOS on Jan. 1st 2020 in Tokyo.

>>> df = azely.compute('Sun', 'ALMA AOS', '2020-01-01', 'Tokyo')

Acceptable formats of each parameter and examples are as follows.

Parameter

Acceptable format

Description

Examples

object

<obj. name>

name of object to be searched

'Sun', 'NGC1068'

<toml>:<obj. name>

user-defined object to be loaded (see below)

'user.toml:M42', 'user:M42' (also valid)

site

'here' (default)

current location (guess by IP address)

<loc. name>

name of location to be searched

'ALMA AOS', 'Tokyo'

<toml>:<loc. name>

user-defined location to be loaded (see below)

'user.toml:ASTE', 'user:ASTE' (also valid)

time

'today' (default)

get one-day time range of today

'now'

get current time

<time>

start time of one-day time range

'2020-01-01', '1/1 12:00', 'Jan. 1st'

<time> to <time>

start and end of time range

'1/1 to 1/3', 'Jan. 1st to Jan. 3rd'

view

'' (default)

use timezone of site

<tz name>

name of timezone database

'Asia/Tokyo', 'UTC'

<loc. name>

name of location from which timezone is identified

same as site’s examples

<toml>:<loc. name>

user-defined location from which timezone is identified

same as site’s examples

Output DataFrame#

The output DataFrame contains az/el expressed in units of degrees and local sidereal time (LST) at site indexed by time in view:

>>> print(df)
                                  az         el             lst
Asia/Tokyo
2020-01-01 00:00:00+09:00  94.820323  68.416756 17:07:59.405556
2020-01-01 00:10:00+09:00  94.333979  70.709575 17:18:01.048298
2020-01-01 00:20:00+09:00  93.856123  73.003864 17:28:02.691044
2020-01-01 00:30:00+09:00  93.388695  75.299436 17:38:04.333786
2020-01-01 00:40:00+09:00  92.935403  77.596109 17:48:05.976529
...                              ...        ...             ...
2020-01-01 23:20:00+09:00  96.711830  59.146249 16:31:49.389513
2020-01-01 23:30:00+09:00  96.185941  61.431823 16:41:51.032256
2020-01-01 23:40:00+09:00  95.664855  63.719668 16:51:52.674998
2020-01-01 23:50:00+09:00  95.147951  66.009577 17:01:54.317740
2020-01-02 00:00:00+09:00  94.634561  68.301349 17:11:55.960483

[145 rows x 3 columns]

Example#

Here is a sample script which will plot one-day elevation of the Sun and candidates of black hole shadow observations at ALMA AOS on Apr. 11th 2017 in UTC.

import azely
import matplotlib.pyplot as plt
plt.style.use('seaborn-whitegrid')

fig, ax = plt.subplots(figsize=(12, 4))

site = 'ALMA AOS'
time = 'Apr. 11th 2017'
view = 'UTC'

for obj in ('Sun', 'Sgr A*', 'M87', 'M104', 'Cen A'):
    df = azely.compute(obj, site, time, view)
    df.el.plot(ax=ax, label=obj)

ax.set_title(f'site: {site}, view: {view}, time: {time}')
ax.set_ylabel('Elevation (deg)')
ax.set_ylim(0, 90)
ax.legend()

multiple-objects.svg

Advanced usage#

This section describes advanced usage of azely by special DataFrame accessor and local TOML files. Note that azely will create a config directory, $XDG_CONFIG_HOME/azely (if the environment variable exists) or ~/.config/azely, after importing azely for the first time. TOML files for configuration (config.toml) and cached information (objects.toml, locations.toml) will be automatically created in it.

Plotting in local sidereal time#

The compute() function does not accept local sidereal time (LST) as view (i.e., view='LST') because LST has no information on year and date. Instead an output DataFrame has in_lst property which provides az/el with a LST index converted from the original time index. For example, the following code will plot elevation of an object in LST:

>>> df.in_lst.el.plot()

In order to use LST values as an index of DataFrame, LST has pseudo dates which start from 1970-01-01. Please ignore them or hide them by using Matplotlib DateFormatter when you plot the result. Here is a sample script which has JST time axis at the bottom and LST axis at the top of a figure, respectively.

import matplotlib.dates as mdates

fig, ax = plt.subplots(figsize=(12, 4))
twin = ax.twiny()

df = azely.compute('Sun', 'Tokyo', '2020-01-01')
df.el.plot(ax=ax, label=df.object.name)
df.in_lst.el.plot(ax=twin, alpha=0)

ax.set_ylabel("Elevation (deg)")
ax.set_ylim(0, 90)
ax.legend()

formatter = mdates.DateFormatter('%H:%M')
twin.xaxis.set_major_formatter(formatter)
fig.autofmt_xdate(rotation=0)

lst-axis.svg

User-defined information#

Azely offers to use user-defined information from a TOML file. Here is a sample TOML file (e.g., user.toml) which has custom object and location informaiton.

# user.toml

[ASTE]
name = "ASTE Telescope"
longitude = "-67.70317915"
latitude = "-22.97163575"
altitude = "0"

[GC]
name = "Galactic center"
frame = "galactic"
longitude = "0deg"
latitude = "0deg"

If it is located in a current directory or in the azely’s config directory, users can use the information like:

>>> df = azely.compute('user:GC', 'user:ASTE', '2020-01-01')

Cached information#

Object and location information obtained from online services is cached to TOML files (objects.toml, locations.toml) in the azely’s config directory with the same format as user-defined files. If a query argument is given with '!' at the end of it, then the cached values are forcibly updated by a new acquisition. This is useful, for example, when users want to update a current location:

>>> df = azely.compute('Sun', 'here!', '2020-01-01')

Customizing default values#

Users can modify default values of the compute() function by editing the azely’s config TOML file (config.toml) in the azely’s config directory like:

# config.toml

[compute]
site = "Tokyo"
time = "now"

Then compute('Sun') becomes equivalent to compute('Sun', 'Tokyo', 'now').