CODEWITHSUNDEEP
pythonpandasastropy
WalkingClouds
WalkingClouds

Reputation: 15

how to convert astropy.table.table.Table file type to pandas.core.frame.DataFrame file type

I have 10 files all of whcih are astropy.table.table.Table file type, all made of same six columns(mjd, filter, flux, flux_error, zp, zpsys) but have different lengths. firstly I want to convert each file to pandas.core.frame.DataFrame file type so that I can add them all into one list and use the pd.concat function to turn all the 10 files into 1 big pandas.core.frame.DataFrame file. I have tried this:

import numpy as np
import pandas as pd
from astropy.table import Table

n=10
li=[]
for i in range(0,n):
    file = "training_data/%s.dat"%i # This way I can call each file automatically
    data = Table.read(file, format="ascii") 
    data = pd.read_table(file) # I convert the file to pandas compatible
    li.append(data) # I add the file into the empty list above
    # now I have my list ready so I compress it into 1 file
all_data = pd.concat(li)

the problem with this method is the all the columns(6 columns) get compressed into 1 column for some reason, this makes it impossible for me to do the rest of the work.

when I check the shape of all_data I get (879, 1). and it looks like this:

all_data.head()

    mjd filter flux flux_error zp zpsys
0   0.0 desg -4.386 4.679 27.5 ab
1   0.011000000005878974 desr -0.5441 2.751 27.5 ab
2   0.027000000001862645 desi 0.4547 4.627 27.5 ab
3   0.043000000005122274 desz -1.047 4.462 27.5 ab
4   13.043000000005122 desg -4.239 4.366 27.5 ab

so how I can I make a file like this but maintain my columns as separate columns?

here is an sample of some my data in file 0:

    mjd     filter  flux   flux_error zp    zpsys
    float64     str4    float64 float64 float64 str2
    0.0       desg      -4.386  4.679   27.5    ab
    0.0110000 desr  -0.5441 2.751   27.5    ab
    0.0270000 desi  0.4547  4.627   27.5    ab
    0.0430000 desz  -1.047  4.462   27.5    ab
    13.043000 desg  -4.239  4.366   27.5    ab
    13.050000 desr  4.695   3.46    27.5    ab
    13.058000 desi  6.291   6.248   27.5    ab
    13.074000 desz  6.412   5.953   27.5    ab
    21.050000 desg  1.588   2.681   27.5    ab
    21.058000 desr  -0.6124 2.171   27.5    ab

Upvotes: 0

Views: 778

Answers (2)

WalkingClouds
WalkingClouds

Reputation: 15

the solution was to include the sep in data = pd.read_table() so that it will keep each column as separate column with specifying the type of sep as "\s+" 

n=10
li=[]
for i in range(0,n):
    file = "training_data/%s.dat"%i # This way I can call each file automatically 
    data = pd.read_table(file, sep="\s+") # I convert the file to pandas compatible
    li.append(data) # I add the file into the empty list above
# now I have my list ready so I compress it into 1 file
all_data = pd.concat(li)

Upvotes: 0

Tom Aldcroft
Tom Aldcroft

Reputation: 2542

It could be that Table.read() is not able to guess the format / delimiter of your data. I'm able to read the included example (data in file 0) using Table.read(file, format='ascii', data_start=2) into a table with 6 columns, but I'm not sure the whitespace is being captured correctly.

I am suspicious that the example data in file 0 is not literally what you are reading, because without the data_start=2 that file will show up with row 1 being "float64 str4 float64 float64 float64 str2".

One thing you can do is try Table.read(file, format='ascii', data_start=2, guess=False).

Upvotes: 1

Related Questions