data-manipulation-utilities 0.0.4__tar.gz → 0.0.8__tar.gz

{data_manipulation_utilities-0.0.4 → data_manipulation_utilities-0.0.8}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: data_manipulation_utilities
-Version: 0.0.4
+Version: 0.0.8
 Description-Content-Type: text/markdown
 Requires-Dist: zfit
 Requires-Dist: logzero
@@ -24,6 +24,33 @@ Requires-Dist: pytest; extra == "dev"
 
 These are tools that can be used for different data analysis tasks.
 
+For LHCb specific tools (not useful for people outside LHCb) see [this](doc/lhcb.md)
+
+# GIT
+
+## Pushing
+
+From the root directory of a version controlled project (i.e. a directory with the `.git` subdirectory) 
+using a `pyproject.toml` file, run:
+
+```bash
+publish
+```
+
+such that:
+
+1. The `pyproject.toml` file is checked and the version of the project is extracted.
+1. If a tag named as the version exists move to the steps below. 
+1. If it does not, make a new tag with the name as the version
+
+Then, for each remote it pushes the tags and the commits. 
+
+*Why?* 
+
+1. Tags should be named as the project's version
+1. As soon as a new version is created, that version needs to be tagged.
+1. In GitHub, one can configure actions to publish projects when the commits are tagged.
+
 # Generic
 
 This section describes generic tools that could not be put in a specific category, but tend to be useful.

{data_manipulation_utilities-0.0.4 → data_manipulation_utilities-0.0.8}/README.md

@@ -2,6 +2,33 @@
 
 These are tools that can be used for different data analysis tasks.
 
+For LHCb specific tools (not useful for people outside LHCb) see [this](doc/lhcb.md)
+
+# GIT
+
+## Pushing
+
+From the root directory of a version controlled project (i.e. a directory with the `.git` subdirectory) 
+using a `pyproject.toml` file, run:
+
+```bash
+publish
+```
+
+such that:
+
+1. The `pyproject.toml` file is checked and the version of the project is extracted.
+1. If a tag named as the version exists move to the steps below. 
+1. If it does not, make a new tag with the name as the version
+
+Then, for each remote it pushes the tags and the commits. 
+
+*Why?* 
+
+1. Tags should be named as the project's version
+1. As soon as a new version is created, that version needs to be tagged.
+1. In GitHub, one can configure actions to publish projects when the commits are tagged.
+
 # Generic
 
 This section describes generic tools that could not be put in a specific category, but tend to be useful.

{data_manipulation_utilities-0.0.4 → data_manipulation_utilities-0.0.8}/pyproject.toml

@@ -1,22 +1,22 @@
 [project]
 name        = 'data_manipulation_utilities'
-version     = '0.0.4'
+version     = '0.0.8'
 readme      = 'README.md'
 dependencies= [
-'zfit', 
+'zfit',
 'logzero',
-'PyYAML', 
-'scipy', 
-'awkward', 
-'tqdm', 
-'joblib', 
-'scikit-learn', 
-'toml', 
-'numpy', 
-'matplotlib', 
-'mplhep', 
+'PyYAML',
+'scipy',
+'awkward',
+'tqdm',
+'joblib',
+'scikit-learn',
+'toml',
+'numpy',
+'matplotlib',
+'mplhep',
 'hist[plot]',
-'polars', 
+'polars',
 'pandas']
 
 [project.optional-dependencies]
@@ -29,6 +29,7 @@ where   = ['src']
 transform_text    ='dmu_scripts.text.transform_text:main'
 coned             ='dmu_scripts.ssh.coned:main'
 check_truth       ='dmu_scripts.physics.check_truth:main'
+update_decinfo    ='dmu_scripts.physics.update_decinfo:main'
 print_trees       ='dmu_scripts.rfile.print_trees:main'
 compare_root_files='dmu_scripts.rfile.compare_root_files:main'
 
@@ -36,3 +37,4 @@ compare_root_files='dmu_scripts.rfile.compare_root_files:main'
 dmu_data=['text/*.txt', 'text/*.toml', 'ml/*/*.yaml', 'plotting/*/*.yaml']
 
 [tool.setuptools]
+script-files=['src/dmu_scripts/git/publish']

{data_manipulation_utilities-0.0.4 → data_manipulation_utilities-0.0.8}/src/data_manipulation_utilities.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: data_manipulation_utilities
-Version: 0.0.4
+Version: 0.0.8
 Description-Content-Type: text/markdown
 Requires-Dist: zfit
 Requires-Dist: logzero
@@ -24,6 +24,33 @@ Requires-Dist: pytest; extra == "dev"
 
 These are tools that can be used for different data analysis tasks.
 
+For LHCb specific tools (not useful for people outside LHCb) see [this](doc/lhcb.md)
+
+# GIT
+
+## Pushing
+
+From the root directory of a version controlled project (i.e. a directory with the `.git` subdirectory) 
+using a `pyproject.toml` file, run:
+
+```bash
+publish
+```
+
+such that:
+
+1. The `pyproject.toml` file is checked and the version of the project is extracted.
+1. If a tag named as the version exists move to the steps below. 
+1. If it does not, make a new tag with the name as the version
+
+Then, for each remote it pushes the tags and the commits. 
+
+*Why?* 
+
+1. Tags should be named as the project's version
+1. As soon as a new version is created, that version needs to be tagged.
+1. In GitHub, one can configure actions to publish projects when the commits are tagged.
+
 # Generic
 
 This section describes generic tools that could not be put in a specific category, but tend to be useful.

{data_manipulation_utilities-0.0.4 → data_manipulation_utilities-0.0.8}/src/data_manipulation_utilities.egg-info/SOURCES.txt

@@ -13,6 +13,7 @@ src/dmu/ml/cv_classifier.py
 src/dmu/ml/cv_predict.py
 src/dmu/ml/train_mva.py
 src/dmu/ml/utilities.py
+src/dmu/physics/utilities.py
 src/dmu/plotting/plotter.py
 src/dmu/plotting/plotter_1d.py
 src/dmu/plotting/plotter_2d.py
@@ -40,7 +41,9 @@ src/dmu_data/text/transform.txt
 src/dmu_data/text/transform_set.toml
 src/dmu_data/text/transform_set.txt
 src/dmu_data/text/transform_trf.txt
+src/dmu_scripts/git/publish
 src/dmu_scripts/physics/check_truth.py
+src/dmu_scripts/physics/update_decinfo.py
 src/dmu_scripts/rfile/compare_root_files.py
 src/dmu_scripts/rfile/print_trees.py
 src/dmu_scripts/ssh/coned.py

{data_manipulation_utilities-0.0.4 → data_manipulation_utilities-0.0.8}/src/data_manipulation_utilities.egg-info/entry_points.txt

@@ -4,3 +4,4 @@ compare_root_files = dmu_scripts.rfile.compare_root_files:main
 coned = dmu_scripts.ssh.coned:main
 print_trees = dmu_scripts.rfile.print_trees:main
 transform_text = dmu_scripts.text.transform_text:main
+update_decinfo = dmu_scripts.physics.update_decinfo:main

data_manipulation_utilities-0.0.8/src/dmu/physics/utilities.py

@@ -0,0 +1,57 @@
+'''
+Module containing utility functions
+'''
+from importlib.resources import files
+from functools           import cache
+
+import yaml
+
+# ---------------------------------
+@cache
+def _get_evt_name() -> dict[str,str]:
+    file_path = files('dmu_data').joinpath('physics/evt_name.yaml')
+    file_path = str(file_path)
+    with open(file_path, encoding='utf-8') as ifile:
+        d_data = yaml.safe_load(ifile)
+
+    return d_data
+# ---------------------------------
+def _format_nickname(nickname : str, style : str) -> str:
+    if style == 'literal':
+        return nickname
+
+    if style != 'safe_1':
+        raise ValueError(f'Invalid style: {style}')
+
+    nickname = nickname.replace('.',   'p')
+    nickname = nickname.replace('-',   'mn')
+    nickname = nickname.replace('+',   'pl')
+    nickname = nickname.replace('=', '_eq_')
+    nickname = nickname.replace(',',    '_')
+
+    return nickname
+# ---------------------------------
+def read_decay_name(event_type : str, style : str = 'safe_1') -> str:
+    '''
+    Takes event type, and style strings, returns nickname of decay as defined in DecFiles package
+
+    Styles:
+
+    literal         : No change is made to nickname
+    safe_1 (default): With following replacements:
+        . -> p
+        = -> _eq_
+        - -> mn
+        + -> pl
+        , -> _
+    '''
+    d_evt_name = _get_evt_name()
+
+    if event_type not in d_evt_name:
+        raise ValueError(f'Event type {event_type} not found')
+
+    value = d_evt_name[event_type]
+    value = _format_nickname(value, style)
+
+    return value
+# ---------------------------------

data_manipulation_utilities-0.0.8/src/dmu_scripts/git/publish

@@ -0,0 +1,88 @@
+#!/usr/bin/env bash
+
+# --------------------------
+display_help()
+{
+    echo "Script meant to:"
+    echo ""
+    echo "1. Check if version in pyproject.toml has been modified"
+    echo "2. If it has create new tag following version name"
+    echo "3. Push to remote "
+}
+# --------------------------
+get_opts()
+{
+    while getopts :hf: option; do 
+        case "${option}" in
+            h)  
+                display_help
+                exit 0
+                ;;  
+           \?)  echo "Invalid option: -${OPTARG}"
+                display_help
+                exit 1
+                ;;  
+            :)  echo "$0: Arguments needed"
+                display_help
+                exit 1
+                ;;  
+        esac
+    done
+}
+# --------------------------
+# Picks up version from pyproject.toml
+get_version()
+{
+    if [[ ! -f pyproject.toml ]];then
+        echo "Cannot find pyproject.toml"
+        exit 1
+    fi
+
+    VERSION_LINE=$(grep version pyproject.toml)
+
+    if [[ $? -ne 0 ]];then
+        ehco "Could not extract version from pyproject.toml"
+        exit 1
+    fi
+
+    if [[ "$VERSION_LINE" =~ .*([0-9]\.[0-9]\.[0-9]).* ]];then
+        VERSION=${BASH_REMATCH[1]}
+        echo "Using version: $VERSION"
+        return
+    fi
+
+    echo "Could not extract version from: $VERSION_LINE"
+    exit 1
+}
+# --------------------------
+create_tag()
+{
+    git tag -n | grep $VERSION
+
+    if [[ $? -eq 0 ]];then
+        echo "Version found among tags, not tagging"
+        return
+    fi
+
+    echo "Version $VERSION not found among tags, creating new tag"
+
+    git tag -a $VERSION
+}
+# --------------------------
+push_all()
+{
+    for REMOTE in $(git remote);do
+        echo "Pushing tags and commits to remote: $REMOTE"
+        git add pyproject.toml
+        git commit -m "Publication commit"
+
+        git push -u $REMOTE HEAD
+        git push    $REMOTE --tags
+    done
+}
+# --------------------------
+get_opts "$@"
+
+get_version
+create_tag
+push_all

data_manipulation_utilities-0.0.8/src/dmu_scripts/physics/update_decinfo.py

@@ -0,0 +1,111 @@
+'''
+Script meant to read information from files in
+
+https://gitlab.cern.ch/lhcb-datapkg/Gen/DecFiles
+
+and store it in current project as data
+'''
+import os
+import re
+import glob
+from dataclasses           import dataclass
+from importlib.resources   import files
+
+import tqdm
+import yaml
+from dmu.logging.log_store import LogStore
+
+log=LogStore.add_logger('dmu_scripts:physics:update_decinfo')
+# ------------------------------
+@dataclass
+class Data:
+    '''
+    Class used to store shared data
+    '''
+    dec_path : str
+    regex    : str = r'#[\s]*[a-zA-Z]+:[\s]*(.*)'
+# ------------------------------
+def _setup() -> None:
+    if 'DECPATH' not in os.environ:
+        raise ValueError('DECPATH, path to root of DecFiles, not found')
+
+    Data.dec_path = os.environ['DECPATH']
+# ------------------------------
+def _line_from_list(file_path : str, contains : str, l_line : list[str]) -> str:
+    l_value = [ line for line in l_line if contains in line ]
+
+    if len(l_value) == 0:
+        log.warning(f'Could not extract {contains} line in: {file_path}')
+        return 'not_found'
+
+    return l_value[0]
+# ------------------------------
+def _val_from_line(file_path : str, line : str) -> str:
+    if line == 'not_found':
+        return line
+
+    mtch = re.match(Data.regex, line)
+    if not mtch:
+        log.warning(f'Cannot extract value from \"{line}\" in file {file_path}')
+        return 'not_found'
+
+    value = mtch.group(1)
+    value = value.replace(' ', '')
+
+    return value
+# ------------------------------
+def _get_evt_name(file_path : str) -> tuple[str,str]:
+    with open(file_path, encoding='utf-8') as ifile:
+        l_line = ifile.read().splitlines()
+
+    evt_line = _line_from_list(file_path, 'EventType', l_line)
+    nam_line = _line_from_list(file_path, 'NickName' , l_line)
+
+    evt_type = _val_from_line(file_path, evt_line)
+    nickname = _val_from_line(file_path, nam_line)
+
+    return evt_type, nickname
+# ------------------------------
+def _read_info() -> dict[str,str]:
+    dec_file_wc = f'{Data.dec_path}/dkfiles/*.dec'
+    l_dec_file  = glob.glob(dec_file_wc)
+    nfiles      = len(l_dec_file)
+    if nfiles == 0:
+        raise ValueError(f'No dec file foudn in {dec_file_wc}')
+
+    log.info(f'Found {nfiles} decay files')
+
+    l_evt_name = [ _get_evt_name(file_path) for file_path in tqdm.tqdm(l_dec_file, ascii=' -') ]
+    d_evt_name = _dict_from_tup_list(l_evt_name)
+
+    return d_evt_name
+# ------------------------------
+def _dict_from_tup_list(l_evt_name : list[tuple[str,str]]) -> dict[str,str]:
+    d_res = {}
+    for key, val in l_evt_name:
+        if key in d_res:
+            old_val = d_res[key]
+            log.warning(f'Key {key} with value {old_val} already found, overriding with {val}')
+
+        d_res[key] = val
+
+    return d_res
+# ------------------------------
+def _dump_info(d_evt_name : dict[str,str]) -> None:
+    yaml_path = files('dmu_data').joinpath('physics/evt_name.yaml')
+    yaml_path = str(yaml_path)
+
+    log.info(f'Saving to: {yaml_path}')
+    with open(yaml_path, 'w', encoding='utf-8') as ofile:
+        yaml.dump(d_evt_name, ofile)
+# ------------------------------
+def main():
+    '''
+    Script starts here
+    '''
+    _setup()
+    d_evt_name = _read_info()
+    _dump_info(d_evt_name)
+# ------------------------------
+if __name__ == '__main__':
+    main()