CODEWITHSUNDEEP
pythonpython-2.7numpy
Alysha
Alysha

Reputation: 13

Comparing two lists of stellar x-y coordinates to find matching objects

I have two .txt files that contain the x and y pixel coordinates of thousands of stars in an image. These two different coordinate lists were the products of different data processing methods, which result in slightly different x and y values for the same object.

File 1 *the id_out is arbitrary

id_out  x_out       y_out         m_out
0       803.6550    907.0910     -8.301
1       700.4570    246.7670     -8.333
2       802.2900    894.2130     -8.344
3       894.6710    780.0040     -8.387

File 2

xcen      ycen      mag           merr
31.662    37.089    22.759        0.387
355.899   37.465    19.969        0.550
103.079   37.000    20.839        0.847
113.500   38.628    20.966        0.796

The objects listed in the .txt files are not organized in a way that allows me to identify the same object in both files. So, I thought for every object in file 1, which has fewer objects than file 2, I would impose a test to find the star match between file 1 and file 2. For every star in file 1, I want to find the star in file 2 that has the closest match to x-y coordinates using the distance formula: distance= sqrt((x1-x2)^2 + (y1-y2)^2) within some tolerance of distance that I can change. Then print onto a master list the x1, y1, x2, y2, m_out, mag, and merr parameters in the file.

Here is the code I have so far, but I am not sure how to arrive at a working solution. 

#/usr/bin/python
import pandas
import numpy as np

xcen_1 = np.genfromtxt('file1.txt', dtype=float, usecols=1)
ycen_1 = np.genfromtxt('file1.txt', dtype=float, usecols=2)
mag1 = np.genfromtxt('file1.txt', dtype=float, usecols=3)

xcen_2 = np.genfromtxt('file2.txt', dtype=float, usecols=0)
ycen_2 = np.genfromtxt('file2.txt', dtype=float, usecols=1)
mag2 = np.genfromtxt('file2.txt', dtype=float, usecols=2)
merr2 = np.genfromtxt('file2.txt', dtype=float, usecols=3)

tolerance=10.0

i=0
file=open('results.txt', 'w+')
file.write("column names")
for i in len(xcen_1):
    dist=np.sqrt((xcen_1[i]-xcen_2[*])^2+(ycen_1[i]-ycen_2[*]^2))
    if dist < tolerance:
        f.write(i, xcen_1, ycen_1, xcen_2, ycen_2, mag1, mag2, merr2)
    else:
        pass
    i=i+1
file.close

The code doesn't work as I don't know how to implement that every star in file 2 must be run through the test, as indicated by the * index (which comes from idl, in which I am more versed). Is there a solution for this logic, as opposed to the thinking in this case:

To compare two independent image coordinate lists with same scale but coordinate-grid having some rotation and shift

Thanks in advance!

Upvotes: 1

Views: 328

Answers (1)

Sameeresque
Sameeresque

Reputation: 2602

You can use pandas Dataframes. Here's how:

import pandas as pd
# files containing the x and y pixel coordinates and other information
df_file1=pd.read_csv('file1.txt',sep='\s+')
df_file2=pd.read_csv('file2.txt',sep='\s+')
join=[]
for i in range(len(df_file1)):
    for j in range(len(df_file2)):
        dis=((df_file1['x_out'][i]-df_file2['xcen'][j])**2+(df_file1['y_out'][i]-df_file2['ycen'][j])**2)**0.5
        if dis<10:
            join.append({'id_out': df_file1['id_out'][i], 'x_out': df_file1['x_out'][i], 'y_out':df_file1['y_out'][i], 
                         'm_out':df_file1['m_out'][i],'xcen':df_file2['xcen'][j],'ycen':df_file2['ycen'][j],
                         'mag':df_file2['mag'][j],'merr':df_file2['merr'][j]})

df_join=pd.DataFrame(join) 
df_join.to_csv('results.txt', sep='\t')

Upvotes: 1

Related Questions