taigapy 2.12.12

No project description provided

taigapy

Library for reading from taiga in python

See here for installing taigr, the library for reading from taiga in R

Token set up

First, you need to get your authorization token so the client library can make requests on your behalf. Go to https://cds.team/taiga/token/ and click on the "Copy" button to copy your token. Paste your token in a file at ~/.taiga/token.

mkdir ~/.taiga/
echo YOUR_TOKEN_HERE > ~/.taiga/token

Installing Taigapy

If you are only using Taigapy and not making modifications to it, run

pip install taigapy

If you are developing Taigapy, check out the repo and run

python setup.py develop

Usage

Main methods

Download

• If you want a file from a dataset to use within Python, use .get method (this will save a file in your cache in feather format)
• If you want all the files from a specific dataset to use within Python, use .get_all method
• If you need a Raw file (for example, plaintext files) or you want to save a file in a specific format, use the .download_to_cache method

Examples:

from taigapy import TaigaClient

tc = TaigaClient() # These two steps could be merged in one with `from taigapy import default_tc as tc`

# The following 6 examples are equivalent, and fetch the file named 'data' from the
# dataset 'Achilles v2.4.6', whose permaname is 'achilles-v2-4-6' and whose current
# version is version 4.

# 1. Fetch a file using dataset ID
df = tc.get(id='022b5ee4df914362945afb8a4dd55d1e')

# 2. Fetch a file using dataset version ID
df = tc.get(id='6d9a6104-e2f8-45cf-9002-df3bcedcb80b')

# 3. Fetch a file using dataset permaname and version
df = tc.get(name='achilles-v2-4-6', version=4)

# 4. Specify the file (optionally if it's the only file, required if there are multiple
#    files)
df = tc.get(name='achilles-v2-4-6', version=4, file='data')

# 5. Fetch a file using dataset permaname and version as the id using the format
#    permaname.version
df = tc.get(id='achilles-v2-4-6.4')

# 6. Specify the file (optionally if it's the only file, required if there are multiple
#    files)
df = tc.get(name='achilles-v2-4-6.4/data')

# If you want the file as a CSV, or want a Raw file, download with download_to_cache
# and get the local file path.
# The options for format are
# - csv, gct, hdf5, tsv for Matrix files
# - csv and tsv for Table files
# - raw for Raw files (this is the default)
raw_path = tc.download_to_cache(name='achilles-v2-4-6', version=4, file='data', format='csv')

Upload

You can also upload data into taiga (see below for available formats). Methods are:

• Create a dataset with create_dataset
• Update a dataset with update_dataset

Examples:

• Create a new dataset in folder public (you can find the folder_id in the url of Taiga web)

from taigapy import TaigaClient

tc = TaigaClient()

# Create a new dataset in public
dataset_id = tc.create_dataset(
    dataset_name='My Dataset Name',
    dataset_description='My Dataset Description',
    upload_file_path_dict={'file_one_path': 'format'},
    folder_id='public'
)

• Update an existing dataset (i.e. create a new dataset version)

from taigapy import TaigaClient

tc = TaigaClient()

# Update a dataset with new files using dataset_id (will use latest version as base)
new_dataset_version_id = tc.update_dataset(
    dataset_id=dataset_id, 
    upload_file_path_dict={'file_updated_or_new_path': 'format'},
)

# Update a dataset using permaname only (will use latest version as base)
new_dataset_version_id = tc.update_dataset(
    dataset_permaname=dataset_permaname,
    upload_file_path_dict={'file_updated_or_new_path': 'format'}
)

# Update a dataset using permaname and version
new_dataset_version_id = tc.update_dataset(
    dataset_permaname=dataset_permaname,
    dataset_version=2,
    upload_file_path_dict={'file_updated_or_new_path': 'format'},
)

# Update a dataset with virtual files (files already on Taiga)
new_dataset_version_id = tc.update_dataset(
    dataset_id=dataset_id,
    add_taiga_ids=['name_in_this_dataset': 'dataset.version/existing_file']
)

# Update a dataset and its description
new_dataset_version_id = tc.update_dataset(
    dataset_id=dataset_id,
    upload_file_path_dict={'file_updated_or_new_path': 'format'},
    description='New description for dataset'
)

# Update a dataset and add a description of changes for this version
new_dataset_version_id = tc.update_dataset(
    dataset_id=dataset_id,
    upload_file_path_dict={'file_updated_or_new_path': 'format'},
    changes_description='Description of changes for this version'
)

# Update a dataset and add all files from the base dataset version as virtual files
new_dataset_version_id = tc.update_dataset(
    dataset_id=dataset_id,
    upload_file_path_dict={'file_new_path': 'format'},
    add_all_existing_files=True
)

Formats available for upload are:

• NumericMatrixCSV
• NumericMatrixTSV
• TableCSV
• TableTSV
• GCT
• Raw

Offline mode

If you are not connected to Taiga but you have a file in your cache, you can get the file from your cache if you call get with id in the dataset_permaname.dataset_version/datafile_name form, or if you provide name, and version, and file.

Running Taigapy via Command line

Run python -m taigapy -h to have an up to date help.

Create a new dataset

python -m taigapy create -n dataset_name -f {'file_path_one': 'format', ...}

Update an existing dataset

python -m taigapy update -p dataset_permaname -v dataset_version -f {'file_path_one': 'format', ...}

Get a dataset from Taiga

python -m taigapy get -p dataset_permaname -v dataset_version -f file_name -t format

[Important] Please choose a format available for this specific file in taiga Web UI

taigaclient

A command-line script, taigaclient, is also available. It is installed with the taigapy package, and it currently supports downloading files to the cache via the fetch command and getting metadata about a dataset via the dataset-meta command.

Usage

usage: taigaclient [-h] [--taiga-url TAIGA_URL] [--data-dir DATA_DIR]
                   {fetch,dataset-meta} ...

optional arguments:
  -h, --help            show this help message and exit
  --taiga-url TAIGA_URL
                        Override default Taiga url (https://cds.team/taiga)
  --data-dir DATA_DIR   Path to where token file lives and cached downloaded
                        files

commands:
  {fetch,dataset-meta}
    fetch               Download a Taiga file into the cache directory
    dataset-meta        Fetch the metadata about a dataset.

fetch

usage: taigaclient fetch [-h] [--name NAME] [--version VERSION] [--file FILE]
                         [--force-fetch] [--quiet] [--force-convert]
                         [--format {raw,feather}]
                         [--write-filename WRITE_FILENAME]
                         [data_file_id]

positional arguments:
  data_file_id          Taiga ID or datafile ID. If not set, NAME must be set

optional arguments:
  -h, --help            show this help message and exit
  --name NAME           Dataset name. Must be set if data_file_id is not set.
  --version VERSION     Dataset version
  --file FILE           Datafile name
  --force-fetch         If set, will bypass local cache and try to redownload
                        from Taiga
  --quiet               If set, do not print progress
  --force-convert       Ask Taiga to convert this file again (Implies --force-
                        fetch)
  --format {raw,feather}
                        Format to store file. If Taiga file is a raw file,
                        choose raw. Otherwise, the default is feather.
  --write-filename WRITE_FILENAME
                        If set, will write the full path and Taiga file type
                        of the cached file to WRITE_FILENAME. Otherwise, will
                        write to stdout

dataset-meta

usage: taigaclient dataset-meta [-h] [--version VERSION]
                                [--write-filename WRITE_FILENAME]
                                dataset_name

positional arguments:
  dataset_name          Dataset name

optional arguments:
  -h, --help            show this help message and exit
  --version VERSION     Dataset version
  --write-filename WRITE_FILENAME
                        Path to JSON file to write metadata

Publish Taigapy on pypi

pip install twine (not to be confused with the interactive fiction software called twine)

Execute: publish_new_taigapy_pypi.sh which will do the following:

1. rm -r dist/
2. python setup.py bdist_wheel --universal
3. twine upload dist/*

More Taigapy information:

Confluence: https://confluence.broadinstitute.org/display/CPDS/Taiga

Running tests:

pytest