Skip to content

A python package with useful methods concernig the FACT-telescope

Notifications You must be signed in to change notification settings

fact-project/pyfact

Repository files navigation

pyfact TravisBuildStatus PyPIStatus

A python package with utils to work with the FACT Imaging Cerenkov Telescope

install with

$ pip install pyfact

This takes automatically care of the dependencies which are installable with pip.

However, if you want to use the GUI Event Viewer you will need to install Tk before you install matplotlib as it depends on the tkagg backend.

functions:

fact includes several functions to convert the times used in fact data to more standard formats and vice versa.

e.g. :

from fact import run2dt

# convert fact fNight format to python datetime object:
date = run2dt("20150101")

Submodules

io

To store pandas dataframes in column-oriented storage into hdf5 files, we created some helpfull wrappers around pandas and h5py:

from fact.io import read_h5py, to_h5py
import pandas as pd

df = pd.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6]})
to_h5py(df, 'test.hdf5', key='events')

print(read_h5py('test.hdf5', key='events'))

plotting

Utils for plotting data into a FACT camera view. Based on matplotlib.

import matplotlib.pyplot as plt
from fact.plotting import camera
from numpy.random import normal

# create some pseudo data with shape (10, 1440):
data = normal(30, 5, (10, 1440))

camera(data[0])
plt.show()

There are also functions to get the camera_geometry from the delivered source file:

from fact.plotting import get_pixel_coords

pixel_x, pixel_y = get_pixel_coords()

factdb

This module contains peewee Models for our factdata MySQL database. These were automatically created by peewee and provide means to query this database in python without writing raw sql queries.

For example, to get the total number of runs take by FACT you can do:

from fact.factdb import connect_database, RunInfo

connect_database()  # this uses the credentials module if no config is given

num_runs = RunInfo.select().count()

A few convenience functions are already implemented. To get a pandas.DataFrame containing the observation time per source and runtype, you can do:

from fact.factdb import connect_database, get_ontime_per_source_and_runtype

connect_database()

num_runs = RunInfo.select().count()
print(get_ontime_by_source_and_runtype())

To download the database and read it to Pandas dataframe without using peewee:

from fact import credentials
import pandas as pd

factDB = credentials.create_factdb_engine()
runInfo = pd.read_sql_table(table_name="RunInfo", con=factDB)

auxservices

Utilities to read in our aux fits files into pandas dataframes.

from fact.auxservices import MagicWeather
from datetime import date

weather = MagicWeather(auxdir='/fact/aux/')

df = weather.read_date(date(2016, 1, 1))