ncar / sunstardb Goto Github PK

the solar-stellar database

Home Page: https://www2.hao.ucar.edu/sunstardb

Python 35.40% Jupyter Notebook 59.89% TSQL 4.70%

sunstardb's Introduction

sunstardb - the solar-stellar database

sunstardb is a publicly accessible database of observations relevant to the study of stellar magnetism. These observations form the basis of the field known as the "solar-stellar connection," which seeks to understand the mechanisms for the production of magnetic field in stars and their dependence on fundamental properties such as mass, luminosity, etc., as well as to understand the place of the Sun's magnetism in this context. sunstardb aggregates and organizes results published in the literature into one place in order to accelerate research efforts in the solar-stellar connection.

sunstardb consists of the following components:

postgreSQL database
python APIs for interactive access
TODO: a web front-end for basic reporting of database contents

More information and announcements may be found at https://www2.hao.ucar.edu/sunstardb

An example can be seen in github by clicking the .ipynb files above, or from this link. (Note that the interactive tables do not work in these links, but do work when running the jupyter notebook server locally by following the installation instructions below.)

Requirements

sunstardb requires the following packages:

Installation

Install the requirements listed above and ensure they are in your python environment.
Download the latest release of sunstardb
In a terminal, change directory to the release downloaded above.
Fetch the configuration file using one of the following options:
- Using wget:
wget --user-agent="Lynx/0 libwww-FM/0" 'https://goo.gl/h1WH2A' -O sunstardb.cfg
- Using curl:
curl -L --user-agent "Lynx/0 libwww-FM/0" 'https://goo.gl/h1WH2A' > sunstardb.cfg
- Or, if neither of the above utilities are installed, using your usual web browser. Go to http://bit.ly/sunstardb_user and save the output to a file named sunstardb.cfg in the top-level folder.
export PYTHONPATH=$PWD:$PYTHONPATH
Run the example notebook:

jupyter notebook 'sunstardb long example.ipynb'

In the jupyter notebook menu, select "Cell -> Run All"

sunstardb's People

Contributors

Stargazers

Watchers

sunstardb's Issues

Bug: Incorrect handling of NULL in results returning astropy Table

For example:

result = db.fetch_data_table('Baliunas1995', ['S_mean', 'activity_class', 'Pcyc', 'Pcyc2'], meta={'Pcyc':'FAPGrade'}, nulls=True

has a dtype:

[('star', 'S12'), ('S_mean', '<f8'), ('activity_class', 'S12'), ('Pcyc', 'O'), ('Pcyc2', 'O'), ('FAPGrade', 'O')]

Which causes all sorts of problems when trying to plot, for example. The ugly workaround is to cast the column post-fetch:

   print result.dtype
   for i, col in enumerate(result.colnames):
        if col in ('Pcyc', 'Pcyc2'):
            dtype[i] = 'float64'
            result[col].fill_value = np.nan
            result[col] = result[col].astype('float64')
        if col in ('FAPGrade'):
            dtype[i] = 'S'
            result[col].mask = [x is None for x in result[col]]
    result = astropy.table.Table(result, dtype=dtype)
    for col in ('Pcyc', 'Pcyc2'):
        result[col].mask = np.isnan(result[col])
    print "Final:", result.dtype

This needs to be handled in fetchall_astropy()

All APIs returning star identifiers should return user-preferred alias

Currently APIs returning stars return the canonical star name, which is set by SIMBAD. Users must build a list of names of their preferred alias (e.g. HD number). For example:

    # Fetch Cycle Data from sunstardb
    result = db.fetch_data_table('Baliunas1995', ['S_mean', 'activity_class', 'Pcyc', 'Pcyc2'], meta={'Pcyc':'FAPGrade'}, nulls=True)

    # Add an HD number column
    hds = []
    hdnums = []
    for star in result['star']:
        if star == 'Sun':
            hd = 'Sun'
            num = 0
        else:
            hd = db.fetch_star_alias(name=star, type='HD')
            num = int(re.match('^HD (\d+)', hd).group(1))
        hds.append(hd)
        hdnums.append(num)
    result['HD'] = hds
    result['Sort'] = hdnums

(Here the 'Sort' column allows the astropy.table.Table to sort the HD numbers in numerical, not lexical, order.)

Instead the user should be able to specify a list of preferred alias types. API example:

    # Fetch Cycle Data from sunstardb
    result = db.fetch_data_table('Baliunas1995', ['S_mean', 'activity_class', 'Pcyc', 'Pcyc2'], meta={'Pcyc':'FAPGrade'}, nulls=True, star_alias=['HD', 'GJ'])

If aliases are not available, the default can be the canonical identifier (star.name in the DB)

Feature: 'providing' option to fetchall_{sources,origins,instruments}

Users should be able to query which sources, origins, or instruments provide a given datatype. The results will be useful for looping queries which build up a dataset from heterogeneous sources. The API could be, e.g.

Prot_sources = db.fetchall_sources(providing='Prot')

Issue connecting to database

I'm getting an error with the command db = sunstardb.database.SunStarDB() which returns OperationalError: could not translate host name "sunstardb.hao.ucar.edu" to address: nodename nor servname provided, or not known.

Trying a ping to sunstardb.hao.ucar.edu looks like the server doesn't exist anymore, is there an up-to-date address to use? Thanks!

fetchall_stars() function

An API should be provided to fetch all stars for which a measurement of a particular datatype exists. API example:

Prot_stars = db.fetchall_stars(with_datatype='Prot')

Other useful options could be from_origin, from_source, from_dataset, from_instrument.

ncar / sunstardb Goto Github PK

sunstardb's Introduction

sunstardb - the solar-stellar database

Requirements

Installation

sunstardb's People

Contributors

Stargazers

Watchers

sunstardb's Issues

Bug: Incorrect handling of NULL in results returning astropy Table

All APIs returning star identifiers should return user-preferred alias

Feature: 'providing' option to fetchall_{sources,origins,instruments}

Issue connecting to database

fetchall_stars() function

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent