Quick Start Tutorial
Getting started with QhX is simple. Follow the steps below to begin analyzing quasar light curves.
Step 1: Installation
First, install the QhX package if you have not already. You can do this using pip:
pip install QhX
Or, if you are installing from the source:
git clone https://github.com/lionandjelka/QhX1.git
cd QhX
pip install .
Step 2: Running Tests
Once you have installed the QhX package, it is important to run the tests to verify that everything is working correctly in your environment. This section explains how to do that.
### 1. Ensure pytest is Installed
To run the tests, you need pytest. You can install it using pip:
pip install pytest
### 2. Check Where the Package is Installed
To confirm where the QhX package is installed, use the following command:
pip show QhX
This command will display the installation path of the QhX package. You can use this path to locate the tests, if necessary.
### 3. Running the Tests
#### Option 1: Running Tests from the Installed Package Directory
If you are in the directory where the package is installed, you can directly run the tests using:
pytest -s QhX/tests/test_integrated.py
#### Option 2: Running Tests from Anywhere
If you are not in the package installation directory, you can specify the full path to the tests. Use the path obtained in step 2, and run:
python -m pytest -s /path/to/your/package/QhX/tests/test_integrated.py
This will run the tests regardless of your current directory, as long as the package is installed.
### 4. Using Verbose Mode
For more detailed test output, you can add the -v flag for verbose mode:
pytest -v -s QhX/tests/test_integrated.py
This will provide more information about each test that is executed.
### Notes
Make sure that you are using the correct Python environment where the QhX package is installed.
You can also run pytest as a Python module to ensure it’s using the correct environment:
python -m pytest -s QhX/tests/test_integrated.py
You should see output similar to the following, indicating that the tests have passed:
====================================================== test session starts ======================================================
platform darwin -- Python 3.10.9, pytest-7.1.2, pluggy-1.0.0
rootdir: /Users/andjelka/Documents/QhX1
plugins: anyio-3.5.0
collected 2 items
QhX/tests/test_integrated.py Running integrated test on simulation of single light curve, and functionalities of modules
for Wavelet matrix coefficients calculations and period and its significance calculation. This may take time about 500-800 seconds...
*** Starting Weighted Wavelet Z-transform ***
Pseudo sample frequency (median) is 0.203
largest tau window is 46.124
8.05 seconds has passed to complete Weighted Wavelet Z-transform
*** Starting Weighted Wavelet Z-transform ***
Pseudo sample frequency (median) is 0.203
largest tau window is 46.124
8.49 seconds has passed to complete Weighted Wavelet Z-transform
.
.
.
results of individual testings of modules for period detection and significance
simulated period in days 100
list of periods in days [100.47726701833716]
upper and lower errors in days [3.3403584182233317] [1.6831563583308196]
number of simulated shuffled light curves 10
experimental significance 1.0
================================================= 1 passed in 114.15s (0:01:54) =================================================
For more advanced users, you may test the logger and parallel solver classes:
os@ubuntu:~/Desktop/qhx_astro/QhX1$ cd QhX/tests; pytest -s test_logger.py
You should see output resembling this:
================================== test session starts ==================================
platform linux -- Python 3.10.11, pytest-8.1.1, pluggy-1.4.0
rootdir: /home/os/Desktop/qhx_astro/QhX1
plugins: anyio-4.3.0
collected 1 item
test_logger.py
Log file created.
Time log found in file.
Output log found in file.
.
================================== 1 passed in 27.56s ===================================
For the parallel solver (note we are in the QhX1 folder again):
os@ubuntu:~/Desktop/qhx_astro/QhX1$ cd QhX/tests; pytest -s test_parallel_solver.py
The output should look similar to this:
================================== test session starts ==================================
platform linux -- Python 3.10.11, pytest-8.1.1, pluggy-1.4.0
rootdir: /home/os/Desktop/qhx_astro/QhX1
plugins: anyio-4.3.0
collected 1 item
test_parallel_solver.py
Log file for 0 created.
Log file for 1 created.
Log file for 2 created.
Log file for 3 created.
Merged results file present.
Merged result file content correct.
.
================================== 1 passed in 23.84s ===================================
Step 3: Importing the Package
Once QhX is installed, and you have confirmed that the tests pass, you can import it into your Python script or interactive session:
import QhX
import numpy as np
import pandas as pd
Step 4: Loading Data
Load your light curve data into QhX. For example importing parquet LSST AGN Data Challange:
from QhX.data_manager import DataManager
data_manager = DataManager()
fs_df = data_manager.load_fs_df('https://zenodo.org/record/6878414/files/ForcedSourceTable.parquet')
fs_gp = data_manager.group_fs_df()
You should see the message like this, indicating that parquet is loaded
Forced source data loaded successfully.
Forced source data grouped successfully.
td_objects=data_manager.load_object_df("https://zenodo.org/record/6878414/files/ObjectTable.parquet")
#Find quasars IDs
setindexqso=td_objects[(td_objects["class"].eq("Qso"))].index
Object data loaded and processed successfully.
##FIND quasars indices and transform to arrays
setindexnew=data_manager.get_qso(setindexqso)
setindexnew=np.array(setindexnew)
df = pd.DataFrame({'objectId': setindexnew})
df.set_index('objectId', inplace=True)
setidnew=df.index
Importing light curve of one object ID=1384142
from QhX.light_curve import get_lctiktok, get_lc22
light_curves_data = get_lc22(data_manager, '1384142', include_errors=False)
Step 5: Analyzing the Light Curve
With the data loaded, you can start analyzing the light curve:
from QhX.calculation import *
from QhX.detection import *
# Ensure to import or define other necessary functions like hybrid2d, periods, same_periods, etc.
from QhX.algorithms.wavelets.wwtz import *
process1_results = process1_new(data_manager, '1384142', ntau=80, ngrid=800, provided_minfq=2000, provided_maxfq=10, include_errors=True)
The output dictionary process1_results contains:
{'objectid': '1384142',
'sampling_i': 45.08568965517243,
'sampling_j': 45.08568965517243,
'period': 446.179587283882,
'upper_error': 26.594480680527795,
'lower_error': 22.513862402711993,
'significance': 0.98,
'label': '0-1'},
{'objectid': '1384142',
'sampling_i': 45.08568965517243,
'sampling_j': 45.08568965517243,
'period': 446.179587283882,
'upper_error': 35.654914195946844,
'lower_error': 17.127186682281263,
'significance': 1.0,
'label': '0-2'},
{'objectid': '1384142',
'sampling_i': 45.08568965517243,
'sampling_j': 45.87666666666669,
'period': 446.179587283882,
'upper_error': 30.9002280462837,
'lower_error': 20.398836831717233,
'significance': 1.0,
'label': '0-3'},
{'objectid': '1384142',
'sampling_i': 45.08568965517243,
'sampling_j': 45.08568965517243,
'period': 472.39444936522017,
'upper_error': 26.594480680527795,
'lower_error': 22.513862402711993,
'significance': 1.0,
'label': '1-2'},
{'objectid': '1384142',
'sampling_i': 45.08568965517243,
'sampling_j': 45.87666666666669,
'period': 472.39444936522017,
'upper_error': 26.594480680527795,
'lower_error': 22.513862402711993,
'significance': 1.0,
'label': '1-3'},
{'objectid': '1384142',
'sampling_i': 45.08568965517243,
'sampling_j': 45.87666666666669,
'period': 308.999613750483,
'upper_error': 22.17397677238779,
'lower_error': 3.746728576973794,
'significance': 0.98,
'label': '1-3'},
{'objectid': '1384142',
'sampling_i': 45.08568965517243,
'sampling_j': 45.87666666666669,
'period': 472.39444936522017,
'upper_error': 30.9002280462837,
'lower_error': 20.398836831717233,
'significance': 0.94,
'label': '2-3'}
Step 6: Viewing Results
Finally, examine the results of your analysis. Important that process1_results will be suplied as a list of dictionaries:
from QhX.output import classify_periods, classify_period
outt=classify_periods([process1_results])
outt['classification'] =outt.apply(classify_period, axis=1)
print(outt)
This will print the detected periods, their errors, significance levels, iou metric, difference among detected periods, flags.
index |
objectid |
m3 |
m4 |
m5 |
m6 |
m7_1 |
m7_2 |
period_diff |
iou |
classification |
|---|---|---|---|---|---|---|---|---|---|---|
0 |
1384142 |
446.179587283882 |
22.5138624027119 |
26.5944806805277 |
0.98 |
0-1 |
0-2 |
0.0 |
1.0 |
poor |
1 |
1384142 |
446.179587283882 |
22.5138624027119 |
26.5944806805277 |
0.98 |
0-1 |
0-3 |
0.0 |
1.0 |
poor |
2 |
1384142 |
446.179587283882 |
22.5138624027119 |
26.5944806805277 |
0.98 |
0-1 |
1-2 |
0.05875 |
0.215 |
poor |
3 |
1384142 |
446.179587283882 |
22.5138624027119 |
26.5944806805277 |
0.98 |
0-1 |
1-3 |
0.05875 |
0.215 |
poor |
4 |
1384142 |
446.179587283882 |
22.5138624027119 |
26.5944806805277 |
0.98 |
0-1 |
1-3 |
0.30745 |
NaN |
NAN |
This table shows an example of the output from the QhX package after analyzing light curve data. The objectid column represents the identifier for the object, while m3 is period. m4, and m5 are upper and lower errors, m6 is significance, to m7_1 and m7_2 columns are the pairs of bands. The period_diff column indicates the difference between detected periods, iou is the intersection over union of the period errors, and the classification column categorizes the reliability of the detected period.
Further Exploration
Now that you’ve had a taste of what QhX can do, explore the documentation to learn more about the available modules and functions. You can also check out the Examples section for more detailed use cases and advanced features.