Astrotools


Useful software
for astrophysicists


Bash, Unix tools
C, C++, Python, PHP

Study PyEphem

python, jupyter notebook

The project presents possibilities of the PyEphem library. This is not a program. It's a Jupyter Notebook document containing Python code. As the main topic it describes How long is a day? The calculations are based on sunrise and sunset moments. There are also considered factors that prolong the day length relative to the night duration. At the end the day length as a function of latitude is shown.

The document can be valuable for students. To perform calculations it uses not only the ephem module, but also the pandas library. Results are presented by means of matplotlib. The lecture note indicates a numerical issue and in a few steps gives its solution showing how to handle with PyEphem carefully.

  • #ephem
  • #day
  • #length
  • #distribution

Unredden stars

python

This program allows to unredden stars on a color-color plane. To do this we need prepare two files. The first file should contain only necessary information about stars with the following columns:

  1. id_star – integer number
  2. Xcolor – float number
  3. Ycolor – float number
  4. Xcolor error – float number
  5. Ycolor error – float number

The second file should contain the theoretical model of stars on that color-color plane (e.g. trace of the main sequence stars) with the simple two-column structure:

  1. Xcolor – float number
  2. Ycolor – float number

Note that it's extremely important that the model must be sorted by increasing temperature. Hence it's usually enough to sort data by the decreasing Xcolor. There are models which can produce hooks, for example theoretical white dwarf sequences, so in such cases we must be more careful. Moreover, we have to estimate a value of the reddening line slope and the parameter R known as the ratio of the total to selective extinction. Having two files and knowing these parameters we can call the script:

$> python unred_stars.py stars.dat model.dat 0.72 3.1

For each star the program prints all intersections of the model and the reddening line passing by the star:

# ID x_ci y_ci x_ci0 y_ci0 E(x_ci) E(y_ci) A               
1 0.0480 -0.4670 -0.1457 -0.6064 0.1937 0.1394 0.6004
1 0.0480 -0.5070 -0.1557 -0.6536 0.2037 0.1466 0.6313
1 0.0480 -0.4270 -0.1357 -0.5593 0.1837 0.1323 0.5694
1 0.0360 -0.4670 -0.1435 -0.5962 0.1795 0.1292 0.5565

Moreover, the script takes into account errors of each color, so for particular star this approach uses nine points instead of one position. For this reason one point can generates a lot of (or zero in some cases) intersections. Using the --min or --max option we can leave only one line for each star with a minimum or maximum value of the estimated extinction.

This image presents a part of the color-color diagram. Gray points indicate background stars toward the Galactic Bulge. The black point mark two stars with known errors. The black line represents model of the main sequence stars. Two straight red lines show the reddening direction. Blue dots points places where the stars should be located if the interstellar extinction will disappear.
Fig. 1. A center part of the U-B vs B-V diagram toward the Galactic Bulge. For each star and its errors (black points with error bars) all intersections (blue dots) with the main sequence model (solid black line) are presented. The red lines are parallel to the reddening line.
  • #reddening
  • #intrinsic
  • #color
  • #excess

Vphasfits

python

This package contains a python module which allows to convert data from the vphas+ project to more convenient formats. The data can be downloaded using the ESO query interface. All data are stored in fits files and are divided into three groups:

  1. catalog
  2. image
  3. source_table

Let's start from importing the module:

>>> from vphasfits import vphaslib

Each image file represents mosaics of 32 sky pawprints. They are enclosed in multi-extension fits files. To get one of a pawprint (here 3) we can use the following function:

>>> vphaslib.pawprint_to_fits("filename.fits", 3)

Source_table file contains the list of stars found on an image (sky/image coordinates, aperture/profile photometry, etc.). To download this data to a text file we can use the function:

>>> vphaslib.srctbl_to_txt("filename.fits", 3)

Catalog file contains the list of stars with the standard photometry in all passbands. To save this data to a text file we should use the following function:

>>> vphaslib.catalog_to_txt("filename.fits")

The data are saved to new files located in the working directory. We can also have an influence on what keys of an image header or which columns should be gathered during the saving process. To do this, please edit any of three lists:

>>> vphaslib.header_keys
>>> vphaslib.srctbl_keys
>>> vphaslib.catalog_keys

Moreover, the package contains three ready-to-use scripts. Each program can be called from the command line and requires arguments (name of file, pawprint number), so anyone has a choice between how to manage the data from the vphasplus project. Unfortunately the last way doesn't allow to edit the above mentioned lists but their default values are well defined.

  • #vphas+
  • #vst
  • #database
  • #fits

Demulos

c, bash, awk, python

Demulos is an acronym for delete multiple objects. This program allows to select isolated objects in dense stellar fields. These objects can be used to determine the PSF model on an image where all stars were found. As an input we need an image in FITS format and a text file containing a list of all stars. The file must have at least the following columns:

  1. id_star – string value
  2. X coordinate – expressed by pixel
  3. Y coordinate – expressed by pixel
  4. brightness – expressed by magnitude
  5. error of brightness – expressed by magnitude

The order of the columns doesn't matter. The structure of the list should be defined inside the demulos.bash file in the == set parameters == section. We just can open the script in any editor and set some variables before use. If the files are prepared we can call the program:

$> bash demulos.bash --list allstars.txt --image image.fits

Additionally one parameter can be also set from the command line:

$> bash demulos.bash --list allstars.txt --image image.fits --diff-mag 3.6

The program creates an initial list of the brightest stars (i = 1, …, N) and for each star from the list calculates a modified distance:

where Mmax denotes the magnitude of the faintest star on the image. If the distance is smaller than a real distance between each of neighboring star, the i-star is not rejected from the initial list. If it is larger, the i-star must be sufficient bright to be not rejected. Thus, if difference between brightness of any neighboring star and considered object (i-star) is smaller than a value of the --diff-mag option, the star is removed from the list. After all the software generates the final list of separated stars in the working directory. This list is stored in a text file which has the same name as the input list with added -demulos suffix.

This picture presents a part of an image of the sky in FITS format. Red circles indicate stars chosen by means of the described software in this article.
Fig. 1. Group of stars in dense field chosen by means of the demulos.bash script.
  • #crowded
  • #fields
  • #PSF model
  • #stars

Astro Quiz

php, twig, html5, css3

This application, based on the MVC architecture, is a little advanced quiz which can be used either on a single computer or on a few machines simultaneously, for example connected through LAN. It is not a good idea to put Astro Quiz on the Internet due to the fact that it doesn't use any popular database. All information are stored in text files and a displayed web page is only an interface between a user and the quiz. The application enables to define own set of questions and scoring.

Fig. 1. Welcome screen.

To run the application we must type localhost into the web browser's address bar. To start Astro Quiz we have to type a name. Each name is validated, particularly whether is duplicated. On the next pages a question with possible answers are displayed. Moreover, an image can be assigned to a question. In each time a sequence of the answers permutates. Let's choose one of four answers and go further.

Fig. 2. Panel with question.

When we achieve the end the application will display our score. Note that all results are saved to a text file which is a simple database located in the database/database file. It stores usernames, collected points and flags marking correctness of each question.

Fig. 3. Results.

As an administrator we can look at results of all users at any time. Let's type localhost/admin.php into the address bar. To go further we have to enter the password which is defined in the astroquiz.cfg file. We can look at tables representing scores and how users answered. Moreover, we can clean the whole database typing the password again.

As mentioned at the beginning we can create own quiz. To do this we have to prepare a text file with questions and images in the case that questions need graphics. The structure of such text file should be following:

Question
Answer-1
Answer-2
Answer-3
Answer-4
Correct Answer
Points
Image name
...

Note that Correct Answer is an integer (1-4). Image name must contain an extension, e.g. .jpg. All these files must be located in the files/ directory. Please see demo files in this localization if you encounter any issues. You can create as many quizzes as you need. The current quiz is defined in the astroquiz.cfg file. In this file you can also set the password for the admin panel and size of images.

  • #quiz
  • #single-choice
  • #test
  • #points

Sigma clipping

bash, awk

This script filters data from a single text file using a particular column. As the title indicates the program uses the σ-clipping algorithm. Let's consider first lines of the input file and focus on seventh column:

# Additional information about file                
1 4326.403 2276.166 0.499 0.259 1 4.1319 0.0318 84
2 4128.059 1329.443 0.040 0.225 0 3.7086 0.0475 37
3 4770.499 2256.309 0.346 0.169 0 4.3550 0.0404 28
4 3497.437 1343.731 0.049 0.190 0 5.9597 0.0191 57

We want to choose points which are centered around the mean value. Moreover, using input parameters we can have an affect on results. To do this, please call the script with some options:

$> bash good_points data.lst 7 1.8 10 80

Note that the name of an input file must have a file extension. This is caused by the fact that results will be stored inside a file with the same name and the .good extension. The above call means that the script analyzes the input file ignoring empty lines and comments. In each iteration it calculates the mean value and the standard deviation, and according to these values the program rejects outlying points. The first argument should be the name of the input file. Only this argument is mandatory. The number 7 indicates a column to study. The standard deviation is scaled by 1.8. The program executes 10 iterations and each iteration must use minimum 80 points. Otherwise it reports that an error has occurred. Moreover, the script outputs some information on the screen:

iteration   average   stand.deviation   rejected
1 4.933370 0.650678 14
2 4.903270 0.542442 16
3 4.852180 0.460161 9
4 4.811320 0.417298 7

All lines which contain remaining points are saved to the data.good file. To see the default values of the optional arguments, please call the program without any arguments. Note that this program can be a great alternative to another programing languages. The Linux default software is sufficient to use the script properly, hence you don't need to install additional packages.

  • #sigma-clipping
  • #rejection
  • #points
  • #selection

Photometric standardization

python

The program enables to make a photometric standardization. It converts instrumental magnitudes to standard values. Only one input file is required. This file must contain a special format related to wavelengths of passbands. Consider four passbands (there is no limitation), e.g. U, B, V, I, then the first five lines of the complete input file with a header may look like (labels from a header are used to sign axes):

# no  U_inst  U_ierr   U_std  U_serr  B_inst  B_ierr   B_std  B_serr  V_inst  V_ierr   V_std  V_serr  I_inst  I_ierr   I_std  I_serr
1 13.1877 0.0065 17.2267 0.0682 11.1404 0.0035 14.4395 0.0079 8.6193 0.0018 12.1166 0.0012 6.9774 0.0032 9.3515 0.0006
4 7.6411 0.0007 11.2698 0.0008 8.1971 0.0010 11.7471 0.0010 7.8598 0.0011 11.2494 0.0007 8.1127 0.0017 10.5117 0.0011
5 13.1125 0.0064 17.0435 0.0604 11.5311 0.0034 14.8228 0.0106 9.3769 0.0017 12.8460 0.0018 8.3673 0.0020 10.7696 0.0012
8 14.9558 0.0232 99.9999 99.9999 12.6868 0.0046 15.8986 0.0289 10.2104 0.0015 13.6863 0.0030 8.6402 0.0013 11.0018 0.0014

The most important thing is that the sequence of consecutive columns must represent passbands with a growing wavelength. Each passband is related with four values: instrumental magnitude, its error, standard magnitude, its error. If any value does't exist, don't worry, it should be masked by 99.9999. Now we can call the program from a command line interface with default arguments:

$> python make_std_phot.py input_file_name output_file_name

For each pair of neighboring passbands the program fits a straight line to a cloud of points with parameters A and B:

Linear standardization equation for two neighboring passbands written in LaTeX.

The fitting, based on the orthogonal distance regression, uses errors from an input file to weight the data. At the end the program generates an output file with converted magnitudes for all stars. Moreover, it produces a log file with parameters and PNG images showing the final fitting for each pair. The strength of this program reveals when we use an interactive mode. Let's call the program again with more options:

$> python make_std_phot.py input_file_name output_file_name -i 5 -s 2.0 -e -v

Now the program makes the fitting iteratively (5 times) removing points that are smaller or larger than m±2σ. After all it displays an interactive window with all points and the final matched line (red). The gray line represents the initial fitting. We can eliminate remaining points (blue dots) just clicking on them. The -e option helps us to identify points with large errors. If you need more information, please use the -h option to display the short manual of the program.

Interactive window of the program. The window displays one of four panels which contains a cloud of points and the fitted straight line.
Fig. 1. Screenshot of the interactive window view.
  • #photometry
  • #standardization
  • #magnitudes
  • #passbands

Interactive CMD and CCD

python

This program has two purposes. The first one is an interactive identification of the same star on different color-magnitude or color-color diagrams, a CMD and CCD respectively. We can make this efficiently. The second is a making images of diagrams additionally marking specific groups of stars. The basic usage requires only one input file with a header and columns with data. The first five lines of an input file may have the following structure (labels from a header are used to sign axes):

# no  U        errU     B        errB     V        errV     I        errI      U-B     errU-B    B-V     errB-V    V-I     errV-I
6 17.4027 0.0084 15.3704 0.0049 13.4930 0.0017 10.2656 0.0024 2.0323 0.0097 1.8774 0.0052 3.2274 0.0029
10 10.8150 0.0009 10.6881 0.0013 10.4042 0.0009 10.0376 0.0028 0.1269 0.0016 0.2839 0.0016 0.3666 0.0029
12 13.9831 0.0012 13.0129 0.0011 11.8005 0.0012 10.5449 0.0027 0.9702 0.0016 1.2124 0.0016 1.2556 0.0030
15 13.8146 0.0013 12.9918 0.0012 11.8464 0.0022 10.5845 0.0033 0.8228 0.0018 1.1454 0.0025 1.2619 0.0040

The input data should be prepared before use. There isn't any mechanism in the program to control a quality and correctness of the data. Let's begin with the simplest call. We can plot as many diagrams as we need (in this description we'll use two diagrams). Each plot will display on a separate window. Assume that we want to look at (U-B vs B-V) and (B vs B-V) diagrams to identify specific stars. Now we have to correlate colors and passbands with particular columns:

$> python photometric_diagrams.py input_file_name --col 12 -10 --col 12 -4

This call means that the program should read the input file and then display two windows with a CCD (10 vs 12) and CMD (4 vs 12). The first argument of the --col option refers to x-coordinates while the second to y-coordinates. The minus value indicates that an axis will be reversed. At least one --col option is mandatory. On the both diagrams the stars are represented by gray points. Clicking on any point changes its color on red – for all windows simultaneously. In this way we can identify an object on different diagrams very quickly. But this isn't the end. If we need more information about the object, we should use the feedback button, which returns an appropriate line from the input file. The feedback is printed to the standard output (console) in the following format:

# object 1
6
17.4027
0.0084
15.3704
0.0049
13.4930
0.0017
10.2656
0.0024
2.0323
0.0097
1.8774
0.0052
3.2274
0.0029

Due to the fact that matplotlib defines an area around the cursor position it is possible to mark more than one object on a crowded diagram. In this case returned information will contain lines of all marked stars separated by # object NR. We can also make a snapshot of the current view of diagrams in any time. There is no limitation of an amount of snapshots. All images are saved as PNG files in the working directory. To make this function more useful we can bring the --grp option into play. If the first column of the input file contains unique numbers of stars, we can group specific objects by different colors. The only thing we need is a simple file with one column which contains numbers of particular objects, for example:

6
10
15

A name of this file should be used as the first argument of the --grp option. The second argument specifies which color to apply marking stars. Assume that we've just created two such files and we want to distinguish them on diagrams. Let's call the program again:

$> python photometric_diagrams.py input_file_name --col 12 -10 --col 12 -4 --grp group_file_name1 green --grp group_file_name2 yellow

More information can be found using the -h option. Note that it's possible to use the program in the wider context. For a globular cluster we can display not only a CMD or CCD diagram but additionally a (RA vs DEC) plot which is a simple 2D map. Only need to prepare the proper input file.

Two separated windows of the program. The windows represent color-color and color-magnitude diagrams. Different colors mark different groups of stars. Each diagram is made of the same stars coming from one input file.
Fig. 1. The CCD and CMD diagrams. Two groups are distinguished by green and yellow color. One selected star is marked by red color.
  • #color
  • #magnitude
  • #diagram
  • #identification

Identification of stars by XY

c

This console program allows to search for common objects in two databases by the XY coordinates. It calculates distances between points and if these are less than the assumed value, it returns matched objects with the smallest separation. Seems to be simple but the program reveals the power when we use its options. Let's prepare two input files where the sample may look like:

# Additional information about file                                      
1 466.569 168.972 7.8349 0.0091 463.696 8. 1.610 -0.021
2 1898.298 934.286 8.0603 0.0140 450.530 7. 3.093 0.134
3 1843.416 1815.547 8.2385 0.0661 471.096 10. 6.909 -0.835
4 134.138 685.132 8.6408 0.0628 448.752 9. 22.962 -0.522

The most important fact is that the first four columns (the last is optional) should have the following structure:

  1. id_star – as an integer number
  2. X coordinate – usually expressed by pixel
  3. Y coordinate – usually expressed by pixel
  4. brightness – usually expressed by magnitude

Moreover, assume that the first file has a one-line header and the second file starts from a two-line header. Let's call the program with a few options:

$> sdb_xy input_file_1 input_file_2 -r 2.8 -h 1 2 -o 0.4 -0.3 -m 1 -s 7

It means that the program will identify objects from two databases using their XY coordinates. The search radius was set to 2.8 px. Because the input files contain headers the -h option ignores one and two first lines in the first and second input file, respectively. The -o option adds an offset to the data. It just transforms coordinates from the first input file adding 0.4 to each X value and -0.3 to each Y value. The -m option defines the output format. In this case the program will print 8 columns: id1, x1, y1, id2, x2, y2, r, mag1-mag2. The -s option sorts the output data by the specified column. Comparing it with the -m option we see that the output will be sorted by r values.

7516   631.044  1849.035    7486   631.353  1848.734   0.0910  -0.5411
2913 392.192 1602.625 2913 392.745 1602.352 0.1554 -0.9238
300 798.892 534.707 3249 799.307 534.222 0.1856 -4.8973
5809 1823.539 1811.843 9182 1824.199 1811.556 0.2603 -0.5848
207 1354.342 690.702 11459 1354.519 690.645 0.3298 -4.7324

For more details please call the program with the --help option. This program is useful when we work with the DAOPHOT package. For example, it helps to control lists of stars in different passbands or to prepare groups of stars to calculate the PSF model.

  • #searching for
  • #points
  • #XY coordinates
  • #common