29ec685ff2
Build pipeline: vsh-ci-dev-9jqlj
Source commit: 5e3f323145
Source message: Fix config error
1614 lines
67 KiB
Bash
Executable File
1614 lines
67 KiB
Bash
Executable File
#!/usr/bin/env bash
|
|
|
|
# onclass fix-integration-tests
|
|
#
|
|
# This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
|
|
# work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
|
|
# Intuitive.
|
|
#
|
|
# The component may contain files which fall under a different license. The
|
|
# authors of this component should specify the license in the header of such
|
|
# files, or include a separate license file detailing the licenses of all included
|
|
# files.
|
|
#
|
|
# Component authors:
|
|
# * Jakub Majercik (author)
|
|
|
|
set -e
|
|
|
|
if [ -z "$VIASH_TEMP" ]; then
|
|
VIASH_TEMP=${VIASH_TEMP:-$VIASH_TMPDIR}
|
|
VIASH_TEMP=${VIASH_TEMP:-$VIASH_TEMPDIR}
|
|
VIASH_TEMP=${VIASH_TEMP:-$VIASH_TMP}
|
|
VIASH_TEMP=${VIASH_TEMP:-$TMPDIR}
|
|
VIASH_TEMP=${VIASH_TEMP:-$TMP}
|
|
VIASH_TEMP=${VIASH_TEMP:-$TEMPDIR}
|
|
VIASH_TEMP=${VIASH_TEMP:-$TEMP}
|
|
VIASH_TEMP=${VIASH_TEMP:-/tmp}
|
|
fi
|
|
|
|
# define helper functions
|
|
# ViashQuote: put quotes around non flag values
|
|
# $1 : unquoted string
|
|
# return : possibly quoted string
|
|
# examples:
|
|
# ViashQuote --foo # returns --foo
|
|
# ViashQuote bar # returns 'bar'
|
|
# Viashquote --foo=bar # returns --foo='bar'
|
|
function ViashQuote {
|
|
if [[ "$1" =~ ^-+[a-zA-Z0-9_\-]+=.+$ ]]; then
|
|
echo "$1" | sed "s#=\(.*\)#='\1'#"
|
|
elif [[ "$1" =~ ^-+[a-zA-Z0-9_\-]+$ ]]; then
|
|
echo "$1"
|
|
else
|
|
echo "'$1'"
|
|
fi
|
|
}
|
|
# ViashRemoveFlags: Remove leading flag
|
|
# $1 : string with a possible leading flag
|
|
# return : string without possible leading flag
|
|
# examples:
|
|
# ViashRemoveFlags --foo=bar # returns bar
|
|
function ViashRemoveFlags {
|
|
echo "$1" | sed 's/^--*[a-zA-Z0-9_\-]*=//'
|
|
}
|
|
# ViashSourceDir: return the path of a bash file, following symlinks
|
|
# usage : ViashSourceDir ${BASH_SOURCE[0]}
|
|
# $1 : Should always be set to ${BASH_SOURCE[0]}
|
|
# returns : The absolute path of the bash file
|
|
function ViashSourceDir {
|
|
local source="$1"
|
|
while [ -h "$source" ]; do
|
|
local dir="$( cd -P "$( dirname "$source" )" >/dev/null 2>&1 && pwd )"
|
|
source="$(readlink "$source")"
|
|
[[ $source != /* ]] && source="$dir/$source"
|
|
done
|
|
cd -P "$( dirname "$source" )" >/dev/null 2>&1 && pwd
|
|
}
|
|
# ViashFindTargetDir: return the path of the '.build.yaml' file, following symlinks
|
|
# usage : ViashFindTargetDir 'ScriptPath'
|
|
# $1 : The location from where to start the upward search
|
|
# returns : The absolute path of the '.build.yaml' file
|
|
function ViashFindTargetDir {
|
|
local source="$1"
|
|
while [[ "$source" != "" && ! -e "$source/.build.yaml" ]]; do
|
|
source=${source%/*}
|
|
done
|
|
echo $source
|
|
}
|
|
# see https://en.wikipedia.org/wiki/Syslog#Severity_level
|
|
VIASH_LOGCODE_EMERGENCY=0
|
|
VIASH_LOGCODE_ALERT=1
|
|
VIASH_LOGCODE_CRITICAL=2
|
|
VIASH_LOGCODE_ERROR=3
|
|
VIASH_LOGCODE_WARNING=4
|
|
VIASH_LOGCODE_NOTICE=5
|
|
VIASH_LOGCODE_INFO=6
|
|
VIASH_LOGCODE_DEBUG=7
|
|
VIASH_VERBOSITY=$VIASH_LOGCODE_NOTICE
|
|
|
|
# ViashLog: Log events depending on the verbosity level
|
|
# usage: ViashLog 1 alert Oh no something went wrong!
|
|
# $1: required verbosity level
|
|
# $2: display tag
|
|
# $3+: messages to display
|
|
# stdout: Your input, prepended by '[$2] '.
|
|
function ViashLog {
|
|
local required_level="$1"
|
|
local display_tag="$2"
|
|
shift 2
|
|
if [ $VIASH_VERBOSITY -ge $required_level ]; then
|
|
>&2 echo "[$display_tag]" "$@"
|
|
fi
|
|
}
|
|
|
|
# ViashEmergency: log events when the system is unstable
|
|
# usage: ViashEmergency Oh no something went wrong.
|
|
# stdout: Your input, prepended by '[emergency] '.
|
|
function ViashEmergency {
|
|
ViashLog $VIASH_LOGCODE_EMERGENCY emergency "$@"
|
|
}
|
|
|
|
# ViashAlert: log events when actions must be taken immediately (e.g. corrupted system database)
|
|
# usage: ViashAlert Oh no something went wrong.
|
|
# stdout: Your input, prepended by '[alert] '.
|
|
function ViashAlert {
|
|
ViashLog $VIASH_LOGCODE_ALERT alert "$@"
|
|
}
|
|
|
|
# ViashCritical: log events when a critical condition occurs
|
|
# usage: ViashCritical Oh no something went wrong.
|
|
# stdout: Your input, prepended by '[critical] '.
|
|
function ViashCritical {
|
|
ViashLog $VIASH_LOGCODE_CRITICAL critical "$@"
|
|
}
|
|
|
|
# ViashError: log events when an error condition occurs
|
|
# usage: ViashError Oh no something went wrong.
|
|
# stdout: Your input, prepended by '[error] '.
|
|
function ViashError {
|
|
ViashLog $VIASH_LOGCODE_ERROR error "$@"
|
|
}
|
|
|
|
# ViashWarning: log potentially abnormal events
|
|
# usage: ViashWarning Something may have gone wrong.
|
|
# stdout: Your input, prepended by '[warning] '.
|
|
function ViashWarning {
|
|
ViashLog $VIASH_LOGCODE_WARNING warning "$@"
|
|
}
|
|
|
|
# ViashNotice: log significant but normal events
|
|
# usage: ViashNotice This just happened.
|
|
# stdout: Your input, prepended by '[notice] '.
|
|
function ViashNotice {
|
|
ViashLog $VIASH_LOGCODE_NOTICE notice "$@"
|
|
}
|
|
|
|
# ViashInfo: log normal events
|
|
# usage: ViashInfo This just happened.
|
|
# stdout: Your input, prepended by '[info] '.
|
|
function ViashInfo {
|
|
ViashLog $VIASH_LOGCODE_INFO info "$@"
|
|
}
|
|
|
|
# ViashDebug: log all events, for debugging purposes
|
|
# usage: ViashDebug This just happened.
|
|
# stdout: Your input, prepended by '[debug] '.
|
|
function ViashDebug {
|
|
ViashLog $VIASH_LOGCODE_DEBUG debug "$@"
|
|
}
|
|
|
|
# find source folder of this component
|
|
VIASH_META_RESOURCES_DIR=`ViashSourceDir ${BASH_SOURCE[0]}`
|
|
|
|
# find the root of the built components & dependencies
|
|
VIASH_TARGET_DIR=`ViashFindTargetDir $VIASH_META_RESOURCES_DIR`
|
|
|
|
# define meta fields
|
|
VIASH_META_NAME="onclass"
|
|
VIASH_META_FUNCTIONALITY_NAME="onclass"
|
|
VIASH_META_EXECUTABLE="$VIASH_META_RESOURCES_DIR/$VIASH_META_NAME"
|
|
VIASH_META_CONFIG="$VIASH_META_RESOURCES_DIR/.config.vsh.yaml"
|
|
VIASH_META_TEMP_DIR="$VIASH_TEMP"
|
|
|
|
|
|
# ViashHelp: Display helpful explanation about this executable
|
|
function ViashHelp {
|
|
echo "onclass fix-integration-tests"
|
|
echo ""
|
|
echo "OnClass is a python package for single-cell cell type annotation. It uses the"
|
|
echo "Cell Ontology to capture the cell type similarity."
|
|
echo "These similarities enable OnClass to annotate cell types that are never seen in"
|
|
echo "the training data."
|
|
echo ""
|
|
echo "Inputs:"
|
|
echo " Input dataset (query) arguments"
|
|
echo ""
|
|
echo " -i, --input"
|
|
echo " type: file, required parameter, file must exist"
|
|
echo " example: input.h5mu"
|
|
echo " The input (query) data to be labeled. Should be a .h5mu file."
|
|
echo ""
|
|
echo " --modality"
|
|
echo " type: string"
|
|
echo " default: rna"
|
|
echo " Which modality to process."
|
|
echo ""
|
|
echo " --input_layer"
|
|
echo " type: string"
|
|
echo " The layer in the input data to be used for cell type annotation if .X is"
|
|
echo " not to be used."
|
|
echo ""
|
|
echo " --cl_nlp_emb_file"
|
|
echo " type: file, required parameter, file must exist"
|
|
echo " The .nlp.emb file with the cell type embeddings."
|
|
echo ""
|
|
echo " --cl_ontology_file"
|
|
echo " type: file, required parameter, file must exist"
|
|
echo " The .ontology file with the cell type ontology."
|
|
echo ""
|
|
echo " --cl_obo_file"
|
|
echo " type: file, required parameter, file must exist"
|
|
echo " The .obo file with the cell type ontology."
|
|
echo ""
|
|
echo " --var_query_gene_names"
|
|
echo " type: string"
|
|
echo " The name of the adata var column in the input data containing gene"
|
|
echo " names; when no gene_name_layer is provided, the var index will be used."
|
|
echo ""
|
|
echo "Reference:"
|
|
echo " Arguments related to the reference dataset."
|
|
echo ""
|
|
echo " --reference"
|
|
echo " type: file, file must exist"
|
|
echo " example: reference.h5mu"
|
|
echo " The reference data to train the CellTypist classifiers on. Only required"
|
|
echo " if a pre-trained --model is not provided."
|
|
echo ""
|
|
echo " --reference_layer"
|
|
echo " type: string"
|
|
echo " The layer in the reference data to be used for cell type annotation if"
|
|
echo " .X is not to be used."
|
|
echo ""
|
|
echo " --reference_obs_target"
|
|
echo " type: string, required parameter"
|
|
echo " example: cell_ontology_class"
|
|
echo " The name of the adata obs column in the reference data containing cell"
|
|
echo " type annotations."
|
|
echo ""
|
|
echo "Outputs:"
|
|
echo " Output arguments."
|
|
echo ""
|
|
echo " --output"
|
|
echo " type: file, output, file must exist"
|
|
echo " example: output.h5mu"
|
|
echo " Output h5mu file."
|
|
echo ""
|
|
echo " --output_compression"
|
|
echo " type: string"
|
|
echo " example: gzip"
|
|
echo " choices: [ gzip, lzf ]"
|
|
echo ""
|
|
echo " --output_obs_predictions"
|
|
echo " type: string"
|
|
echo " default: onclass_pred"
|
|
echo " In which \`.obs\` slots to store the predicted information."
|
|
echo ""
|
|
echo " --output_obs_probability"
|
|
echo " type: string"
|
|
echo " default: onclass_prob"
|
|
echo " In which \`.obs\` slots to store the probability of the predictions."
|
|
echo ""
|
|
echo "Model arguments:"
|
|
echo " Model arguments"
|
|
echo ""
|
|
echo " --model"
|
|
echo " type: string"
|
|
echo " \"Pretrained model path without a file extension. If not provided, the"
|
|
echo " model will be trained"
|
|
echo " on the reference data and --reference should be provided. The path"
|
|
echo " namespace should contain:"
|
|
echo " - a .npz or .pkl file"
|
|
echo " - a .data file"
|
|
echo " - a .meta file"
|
|
echo " - a .index file"
|
|
echo " e.g. /path/to/model/pretrained_model_target1 as saved by OnClass.\""
|
|
echo ""
|
|
echo " --max_iter"
|
|
echo " type: integer"
|
|
echo " default: 30"
|
|
echo " Maximum number of iterations for training the model."
|
|
}
|
|
|
|
# initialise variables
|
|
VIASH_MODE='run'
|
|
VIASH_ENGINE_ID='docker'
|
|
|
|
######## Helper functions for setting up Docker images for viash ########
|
|
# expects: ViashDockerBuild
|
|
|
|
# ViashDockerInstallationCheck: check whether Docker is installed correctly
|
|
#
|
|
# examples:
|
|
# ViashDockerInstallationCheck
|
|
function ViashDockerInstallationCheck {
|
|
ViashDebug "Checking whether Docker is installed"
|
|
if [ ! command -v docker &> /dev/null ]; then
|
|
ViashCritical "Docker doesn't seem to be installed. See 'https://docs.docker.com/get-docker/' for instructions."
|
|
exit 1
|
|
fi
|
|
|
|
ViashDebug "Checking whether the Docker daemon is running"
|
|
local save=$-; set +e
|
|
local docker_version=$(docker version --format '{{.Client.APIVersion}}' 2> /dev/null)
|
|
local out=$?
|
|
[[ $save =~ e ]] && set -e
|
|
if [ $out -ne 0 ]; then
|
|
ViashCritical "Docker daemon does not seem to be running. Try one of the following:"
|
|
ViashCritical "- Try running 'dockerd' in the command line"
|
|
ViashCritical "- See https://docs.docker.com/config/daemon/"
|
|
exit 1
|
|
fi
|
|
}
|
|
|
|
# ViashDockerRemoteTagCheck: check whether a Docker image is available
|
|
# on a remote. Assumes `docker login` has been performed, if relevant.
|
|
#
|
|
# $1 : image identifier with format `[registry/]image[:tag]`
|
|
# exit code $? : whether or not the image was found
|
|
# examples:
|
|
# ViashDockerRemoteTagCheck python:latest
|
|
# echo $? # returns '0'
|
|
# ViashDockerRemoteTagCheck sdaizudceahifu
|
|
# echo $? # returns '1'
|
|
function ViashDockerRemoteTagCheck {
|
|
docker manifest inspect $1 > /dev/null 2> /dev/null
|
|
}
|
|
|
|
# ViashDockerLocalTagCheck: check whether a Docker image is available locally
|
|
#
|
|
# $1 : image identifier with format `[registry/]image[:tag]`
|
|
# exit code $? : whether or not the image was found
|
|
# examples:
|
|
# docker pull python:latest
|
|
# ViashDockerLocalTagCheck python:latest
|
|
# echo $? # returns '0'
|
|
# ViashDockerLocalTagCheck sdaizudceahifu
|
|
# echo $? # returns '1'
|
|
function ViashDockerLocalTagCheck {
|
|
[ -n "$(docker images -q $1)" ]
|
|
}
|
|
|
|
# ViashDockerPull: pull a Docker image
|
|
#
|
|
# $1 : image identifier with format `[registry/]image[:tag]`
|
|
# exit code $? : whether or not the image was found
|
|
# examples:
|
|
# ViashDockerPull python:latest
|
|
# echo $? # returns '0'
|
|
# ViashDockerPull sdaizudceahifu
|
|
# echo $? # returns '1'
|
|
function ViashDockerPull {
|
|
ViashNotice "Checking if Docker image is available at '$1'"
|
|
if [ $VIASH_VERBOSITY -ge $VIASH_LOGCODE_INFO ]; then
|
|
docker pull $1 && return 0 || return 1
|
|
else
|
|
local save=$-; set +e
|
|
docker pull $1 2> /dev/null > /dev/null
|
|
local out=$?
|
|
[[ $save =~ e ]] && set -e
|
|
if [ $out -ne 0 ]; then
|
|
ViashWarning "Could not pull from '$1'. Docker image doesn't exist or is not accessible."
|
|
fi
|
|
return $out
|
|
fi
|
|
}
|
|
|
|
# ViashDockerPush: push a Docker image
|
|
#
|
|
# $1 : image identifier with format `[registry/]image[:tag]`
|
|
# exit code $? : whether or not the image was found
|
|
# examples:
|
|
# ViashDockerPush python:latest
|
|
# echo $? # returns '0'
|
|
# ViashDockerPush sdaizudceahifu
|
|
# echo $? # returns '1'
|
|
function ViashDockerPush {
|
|
ViashNotice "Pushing image to '$1'"
|
|
local save=$-; set +e
|
|
local out
|
|
if [ $VIASH_VERBOSITY -ge $VIASH_LOGCODE_INFO ]; then
|
|
docker push $1
|
|
out=$?
|
|
else
|
|
docker push $1 2> /dev/null > /dev/null
|
|
out=$?
|
|
fi
|
|
[[ $save =~ e ]] && set -e
|
|
if [ $out -eq 0 ]; then
|
|
ViashNotice "Container '$1' push succeeded."
|
|
else
|
|
ViashError "Container '$1' push errored. You might not be logged in or have the necessary permissions."
|
|
fi
|
|
return $out
|
|
}
|
|
|
|
# ViashDockerPullElseBuild: pull a Docker image, else build it
|
|
#
|
|
# $1 : image identifier with format `[registry/]image[:tag]`
|
|
# ViashDockerBuild : a Bash function which builds a docker image, takes image identifier as argument.
|
|
# examples:
|
|
# ViashDockerPullElseBuild mynewcomponent
|
|
function ViashDockerPullElseBuild {
|
|
local save=$-; set +e
|
|
ViashDockerPull $1
|
|
local out=$?
|
|
[[ $save =~ e ]] && set -e
|
|
if [ $out -ne 0 ]; then
|
|
ViashDockerBuild $@
|
|
fi
|
|
}
|
|
|
|
# ViashDockerSetup: create a Docker image, according to specified docker setup strategy
|
|
#
|
|
# $1 : image identifier with format `[registry/]image[:tag]`
|
|
# $2 : docker setup strategy, see DockerSetupStrategy.scala
|
|
# examples:
|
|
# ViashDockerSetup mynewcomponent alwaysbuild
|
|
function ViashDockerSetup {
|
|
local image_id="$1"
|
|
local setup_strategy="$2"
|
|
if [ "$setup_strategy" == "alwaysbuild" -o "$setup_strategy" == "build" -o "$setup_strategy" == "b" ]; then
|
|
ViashDockerBuild $image_id --no-cache $(ViashDockerBuildArgs "$engine_id")
|
|
elif [ "$setup_strategy" == "alwayspull" -o "$setup_strategy" == "pull" -o "$setup_strategy" == "p" ]; then
|
|
ViashDockerPull $image_id
|
|
elif [ "$setup_strategy" == "alwayspullelsebuild" -o "$setup_strategy" == "pullelsebuild" ]; then
|
|
ViashDockerPullElseBuild $image_id --no-cache $(ViashDockerBuildArgs "$engine_id")
|
|
elif [ "$setup_strategy" == "alwayspullelsecachedbuild" -o "$setup_strategy" == "pullelsecachedbuild" ]; then
|
|
ViashDockerPullElseBuild $image_id $(ViashDockerBuildArgs "$engine_id")
|
|
elif [ "$setup_strategy" == "alwayscachedbuild" -o "$setup_strategy" == "cachedbuild" -o "$setup_strategy" == "cb" ]; then
|
|
ViashDockerBuild $image_id $(ViashDockerBuildArgs "$engine_id")
|
|
elif [[ "$setup_strategy" =~ ^ifneedbe ]]; then
|
|
local save=$-; set +e
|
|
ViashDockerLocalTagCheck $image_id
|
|
local outCheck=$?
|
|
[[ $save =~ e ]] && set -e
|
|
if [ $outCheck -eq 0 ]; then
|
|
ViashInfo "Image $image_id already exists"
|
|
elif [ "$setup_strategy" == "ifneedbebuild" ]; then
|
|
ViashDockerBuild $image_id --no-cache $(ViashDockerBuildArgs "$engine_id")
|
|
elif [ "$setup_strategy" == "ifneedbecachedbuild" ]; then
|
|
ViashDockerBuild $image_id $(ViashDockerBuildArgs "$engine_id")
|
|
elif [ "$setup_strategy" == "ifneedbepull" ]; then
|
|
ViashDockerPull $image_id
|
|
elif [ "$setup_strategy" == "ifneedbepullelsebuild" ]; then
|
|
ViashDockerPullElseBuild $image_id --no-cache $(ViashDockerBuildArgs "$engine_id")
|
|
elif [ "$setup_strategy" == "ifneedbepullelsecachedbuild" ]; then
|
|
ViashDockerPullElseBuild $image_id $(ViashDockerBuildArgs "$engine_id")
|
|
else
|
|
ViashError "Unrecognised Docker strategy: $setup_strategy"
|
|
exit 1
|
|
fi
|
|
elif [ "$setup_strategy" == "push" -o "$setup_strategy" == "forcepush" -o "$setup_strategy" == "alwayspush" ]; then
|
|
ViashDockerPush "$image_id"
|
|
elif [ "$setup_strategy" == "pushifnotpresent" -o "$setup_strategy" == "gentlepush" -o "$setup_strategy" == "maybepush" ]; then
|
|
local save=$-; set +e
|
|
ViashDockerRemoteTagCheck $image_id
|
|
local outCheck=$?
|
|
[[ $save =~ e ]] && set -e
|
|
if [ $outCheck -eq 0 ]; then
|
|
ViashNotice "Container '$image_id' exists, doing nothing."
|
|
else
|
|
ViashNotice "Container '$image_id' does not yet exist."
|
|
ViashDockerPush "$image_id"
|
|
fi
|
|
elif [ "$setup_strategy" == "donothing" -o "$setup_strategy" == "meh" ]; then
|
|
ViashNotice "Skipping setup."
|
|
else
|
|
ViashError "Unrecognised Docker strategy: $setup_strategy"
|
|
exit 1
|
|
fi
|
|
}
|
|
|
|
# ViashDockerCheckCommands: Check whether a docker container has the required commands
|
|
#
|
|
# $1 : image identifier with format `[registry/]image[:tag]`
|
|
# $@ : commands to verify being present
|
|
# examples:
|
|
# ViashDockerCheckCommands bash:4.0 bash ps foo
|
|
function ViashDockerCheckCommands {
|
|
local image_id="$1"
|
|
shift 1
|
|
local commands="$@"
|
|
local save=$-; set +e
|
|
local missing # mark 'missing' as local in advance, otherwise the exit code of the command will be missing and always be '0'
|
|
missing=$(docker run --rm --entrypoint=sh "$image_id" -c "for command in $commands; do command -v \$command >/dev/null 2>&1; if [ \$? -ne 0 ]; then echo \$command; exit 1; fi; done")
|
|
local outCheck=$?
|
|
[[ $save =~ e ]] && set -e
|
|
if [ $outCheck -ne 0 ]; then
|
|
ViashError "Docker container '$image_id' does not contain command '$missing'."
|
|
exit 1
|
|
fi
|
|
}
|
|
|
|
# ViashDockerBuild: build a docker image
|
|
# $1 : image identifier with format `[registry/]image[:tag]`
|
|
# $... : additional arguments to pass to docker build
|
|
# $VIASH_META_TEMP_DIR : temporary directory to store dockerfile & optional resources in
|
|
# $VIASH_META_NAME : name of the component
|
|
# $VIASH_META_RESOURCES_DIR : directory containing the resources
|
|
# $VIASH_VERBOSITY : verbosity level
|
|
# exit code $? : whether or not the image was built successfully
|
|
function ViashDockerBuild {
|
|
local image_id="$1"
|
|
shift 1
|
|
|
|
# create temporary directory to store dockerfile & optional resources in
|
|
local tmpdir=$(mktemp -d "$VIASH_META_TEMP_DIR/dockerbuild-$VIASH_META_NAME-XXXXXX")
|
|
local dockerfile="$tmpdir/Dockerfile"
|
|
function clean_up {
|
|
rm -rf "$tmpdir"
|
|
}
|
|
trap clean_up EXIT
|
|
|
|
# store dockerfile and resources
|
|
ViashDockerfile "$VIASH_ENGINE_ID" > "$dockerfile"
|
|
|
|
# generate the build command
|
|
local docker_build_cmd="docker build -t '$image_id' $@ '$VIASH_META_RESOURCES_DIR' -f '$dockerfile'"
|
|
|
|
# build the container
|
|
ViashNotice "Building container '$image_id' with Dockerfile"
|
|
ViashInfo "$docker_build_cmd"
|
|
local save=$-; set +e
|
|
if [ $VIASH_VERBOSITY -ge $VIASH_LOGCODE_INFO ]; then
|
|
eval $docker_build_cmd
|
|
else
|
|
eval $docker_build_cmd &> "$tmpdir/docker_build.log"
|
|
fi
|
|
|
|
# check exit code
|
|
local out=$?
|
|
[[ $save =~ e ]] && set -e
|
|
if [ $out -ne 0 ]; then
|
|
ViashError "Error occurred while building container '$image_id'"
|
|
if [ $VIASH_VERBOSITY -lt $VIASH_LOGCODE_INFO ]; then
|
|
ViashError "Transcript: --------------------------------"
|
|
cat "$tmpdir/docker_build.log"
|
|
ViashError "End of transcript --------------------------"
|
|
fi
|
|
exit 1
|
|
fi
|
|
}
|
|
|
|
######## End of helper functions for setting up Docker images for viash ########
|
|
|
|
# ViashDockerFile: print the dockerfile to stdout
|
|
# $1 : engine identifier
|
|
# return : dockerfile required to run this component
|
|
# examples:
|
|
# ViashDockerFile
|
|
function ViashDockerfile {
|
|
local engine_id="$1"
|
|
|
|
if [[ "$engine_id" == "docker" ]]; then
|
|
cat << 'VIASHDOCKER'
|
|
FROM python:3.8
|
|
ENTRYPOINT []
|
|
RUN pip install --upgrade pip && \
|
|
pip install --upgrade --no-cache-dir "scikit-learn==0.24.0" "OnClass==1.2" "tensorflow==2.13.1" "obonet==1.1.0" "mudata"
|
|
|
|
LABEL org.opencontainers.image.authors="Jakub Majercik"
|
|
LABEL org.opencontainers.image.description="Companion container for running component annotate onclass"
|
|
LABEL org.opencontainers.image.created="2024-11-18T09:22:55Z"
|
|
LABEL org.opencontainers.image.source="https://github.com/openpipelines-bio/openpipeline"
|
|
LABEL org.opencontainers.image.revision="5e3f3231456facaf14da57fa563d4de58492ca82"
|
|
LABEL org.opencontainers.image.version="fix-integration-tests"
|
|
|
|
VIASHDOCKER
|
|
fi
|
|
}
|
|
|
|
# ViashDockerBuildArgs: return the arguments to pass to docker build
|
|
# $1 : engine identifier
|
|
# return : arguments to pass to docker build
|
|
function ViashDockerBuildArgs {
|
|
local engine_id="$1"
|
|
|
|
if [[ "$engine_id" == "docker" ]]; then
|
|
echo ""
|
|
fi
|
|
}
|
|
|
|
# ViashAbsolutePath: generate absolute path from relative path
|
|
# borrowed from https://stackoverflow.com/a/21951256
|
|
# $1 : relative filename
|
|
# return : absolute path
|
|
# examples:
|
|
# ViashAbsolutePath some_file.txt # returns /path/to/some_file.txt
|
|
# ViashAbsolutePath /foo/bar/.. # returns /foo
|
|
function ViashAbsolutePath {
|
|
local thePath
|
|
local parr
|
|
local outp
|
|
local len
|
|
if [[ ! "$1" =~ ^/ ]]; then
|
|
thePath="$PWD/$1"
|
|
else
|
|
thePath="$1"
|
|
fi
|
|
echo "$thePath" | (
|
|
IFS=/
|
|
read -a parr
|
|
declare -a outp
|
|
for i in "${parr[@]}"; do
|
|
case "$i" in
|
|
''|.) continue ;;
|
|
..)
|
|
len=${#outp[@]}
|
|
if ((len==0)); then
|
|
continue
|
|
else
|
|
unset outp[$((len-1))]
|
|
fi
|
|
;;
|
|
*)
|
|
len=${#outp[@]}
|
|
outp[$len]="$i"
|
|
;;
|
|
esac
|
|
done
|
|
echo /"${outp[*]}"
|
|
)
|
|
}
|
|
# ViashDockerAutodetectMount: auto configuring docker mounts from parameters
|
|
# $1 : The parameter value
|
|
# returns : New parameter
|
|
# $VIASH_DIRECTORY_MOUNTS : Added another parameter to be passed to docker
|
|
# $VIASH_DOCKER_AUTOMOUNT_PREFIX : The prefix to be used for the automounts
|
|
# examples:
|
|
# ViashDockerAutodetectMount /path/to/bar # returns '/viash_automount/path/to/bar'
|
|
# ViashDockerAutodetectMountArg /path/to/bar # returns '--volume="/path/to:/viash_automount/path/to"'
|
|
function ViashDockerAutodetectMount {
|
|
local abs_path=$(ViashAbsolutePath "$1")
|
|
local mount_source
|
|
local base_name
|
|
if [ -d "$abs_path" ]; then
|
|
mount_source="$abs_path"
|
|
base_name=""
|
|
else
|
|
mount_source=`dirname "$abs_path"`
|
|
base_name=`basename "$abs_path"`
|
|
fi
|
|
local mount_target="$VIASH_DOCKER_AUTOMOUNT_PREFIX$mount_source"
|
|
if [ -z "$base_name" ]; then
|
|
echo "$mount_target"
|
|
else
|
|
echo "$mount_target/$base_name"
|
|
fi
|
|
}
|
|
function ViashDockerAutodetectMountArg {
|
|
local abs_path=$(ViashAbsolutePath "$1")
|
|
local mount_source
|
|
local base_name
|
|
if [ -d "$abs_path" ]; then
|
|
mount_source="$abs_path"
|
|
base_name=""
|
|
else
|
|
mount_source=`dirname "$abs_path"`
|
|
base_name=`basename "$abs_path"`
|
|
fi
|
|
local mount_target="$VIASH_DOCKER_AUTOMOUNT_PREFIX$mount_source"
|
|
ViashDebug "ViashDockerAutodetectMountArg $1 -> $mount_source -> $mount_target"
|
|
echo "--volume=\"$mount_source:$mount_target\""
|
|
}
|
|
function ViashDockerStripAutomount {
|
|
local abs_path=$(ViashAbsolutePath "$1")
|
|
echo "${abs_path#$VIASH_DOCKER_AUTOMOUNT_PREFIX}"
|
|
}
|
|
# initialise variables
|
|
VIASH_DIRECTORY_MOUNTS=()
|
|
|
|
# configure default docker automount prefix if it is unset
|
|
if [ -z "${VIASH_DOCKER_AUTOMOUNT_PREFIX+x}" ]; then
|
|
VIASH_DOCKER_AUTOMOUNT_PREFIX="/viash_automount"
|
|
fi
|
|
|
|
# initialise docker variables
|
|
VIASH_DOCKER_RUN_ARGS=(-i --rm)
|
|
|
|
# initialise array
|
|
VIASH_POSITIONAL_ARGS=''
|
|
|
|
while [[ $# -gt 0 ]]; do
|
|
case "$1" in
|
|
-h|--help)
|
|
ViashHelp
|
|
exit
|
|
;;
|
|
---v|---verbose)
|
|
let "VIASH_VERBOSITY=VIASH_VERBOSITY+1"
|
|
shift 1
|
|
;;
|
|
---verbosity)
|
|
VIASH_VERBOSITY="$2"
|
|
shift 2
|
|
;;
|
|
---verbosity=*)
|
|
VIASH_VERBOSITY="$(ViashRemoveFlags "$1")"
|
|
shift 1
|
|
;;
|
|
--version)
|
|
echo "onclass fix-integration-tests"
|
|
exit
|
|
;;
|
|
--input)
|
|
[ -n "$VIASH_PAR_INPUT" ] && ViashError Bad arguments for option \'--input\': \'$VIASH_PAR_INPUT\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_INPUT="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --input. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--input=*)
|
|
[ -n "$VIASH_PAR_INPUT" ] && ViashError Bad arguments for option \'--input=*\': \'$VIASH_PAR_INPUT\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_INPUT=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
-i)
|
|
[ -n "$VIASH_PAR_INPUT" ] && ViashError Bad arguments for option \'-i\': \'$VIASH_PAR_INPUT\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_INPUT="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to -i. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--modality)
|
|
[ -n "$VIASH_PAR_MODALITY" ] && ViashError Bad arguments for option \'--modality\': \'$VIASH_PAR_MODALITY\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_MODALITY="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --modality. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--modality=*)
|
|
[ -n "$VIASH_PAR_MODALITY" ] && ViashError Bad arguments for option \'--modality=*\': \'$VIASH_PAR_MODALITY\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_MODALITY=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--input_layer)
|
|
[ -n "$VIASH_PAR_INPUT_LAYER" ] && ViashError Bad arguments for option \'--input_layer\': \'$VIASH_PAR_INPUT_LAYER\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_INPUT_LAYER="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --input_layer. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--input_layer=*)
|
|
[ -n "$VIASH_PAR_INPUT_LAYER" ] && ViashError Bad arguments for option \'--input_layer=*\': \'$VIASH_PAR_INPUT_LAYER\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_INPUT_LAYER=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--cl_nlp_emb_file)
|
|
[ -n "$VIASH_PAR_CL_NLP_EMB_FILE" ] && ViashError Bad arguments for option \'--cl_nlp_emb_file\': \'$VIASH_PAR_CL_NLP_EMB_FILE\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_CL_NLP_EMB_FILE="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --cl_nlp_emb_file. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--cl_nlp_emb_file=*)
|
|
[ -n "$VIASH_PAR_CL_NLP_EMB_FILE" ] && ViashError Bad arguments for option \'--cl_nlp_emb_file=*\': \'$VIASH_PAR_CL_NLP_EMB_FILE\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_CL_NLP_EMB_FILE=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--cl_ontology_file)
|
|
[ -n "$VIASH_PAR_CL_ONTOLOGY_FILE" ] && ViashError Bad arguments for option \'--cl_ontology_file\': \'$VIASH_PAR_CL_ONTOLOGY_FILE\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_CL_ONTOLOGY_FILE="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --cl_ontology_file. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--cl_ontology_file=*)
|
|
[ -n "$VIASH_PAR_CL_ONTOLOGY_FILE" ] && ViashError Bad arguments for option \'--cl_ontology_file=*\': \'$VIASH_PAR_CL_ONTOLOGY_FILE\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_CL_ONTOLOGY_FILE=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--cl_obo_file)
|
|
[ -n "$VIASH_PAR_CL_OBO_FILE" ] && ViashError Bad arguments for option \'--cl_obo_file\': \'$VIASH_PAR_CL_OBO_FILE\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_CL_OBO_FILE="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --cl_obo_file. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--cl_obo_file=*)
|
|
[ -n "$VIASH_PAR_CL_OBO_FILE" ] && ViashError Bad arguments for option \'--cl_obo_file=*\': \'$VIASH_PAR_CL_OBO_FILE\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_CL_OBO_FILE=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--var_query_gene_names)
|
|
[ -n "$VIASH_PAR_VAR_QUERY_GENE_NAMES" ] && ViashError Bad arguments for option \'--var_query_gene_names\': \'$VIASH_PAR_VAR_QUERY_GENE_NAMES\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_VAR_QUERY_GENE_NAMES="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --var_query_gene_names. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--var_query_gene_names=*)
|
|
[ -n "$VIASH_PAR_VAR_QUERY_GENE_NAMES" ] && ViashError Bad arguments for option \'--var_query_gene_names=*\': \'$VIASH_PAR_VAR_QUERY_GENE_NAMES\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_VAR_QUERY_GENE_NAMES=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--reference)
|
|
[ -n "$VIASH_PAR_REFERENCE" ] && ViashError Bad arguments for option \'--reference\': \'$VIASH_PAR_REFERENCE\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_REFERENCE="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --reference. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--reference=*)
|
|
[ -n "$VIASH_PAR_REFERENCE" ] && ViashError Bad arguments for option \'--reference=*\': \'$VIASH_PAR_REFERENCE\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_REFERENCE=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--reference_layer)
|
|
[ -n "$VIASH_PAR_REFERENCE_LAYER" ] && ViashError Bad arguments for option \'--reference_layer\': \'$VIASH_PAR_REFERENCE_LAYER\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_REFERENCE_LAYER="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --reference_layer. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--reference_layer=*)
|
|
[ -n "$VIASH_PAR_REFERENCE_LAYER" ] && ViashError Bad arguments for option \'--reference_layer=*\': \'$VIASH_PAR_REFERENCE_LAYER\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_REFERENCE_LAYER=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--reference_obs_target)
|
|
[ -n "$VIASH_PAR_REFERENCE_OBS_TARGET" ] && ViashError Bad arguments for option \'--reference_obs_target\': \'$VIASH_PAR_REFERENCE_OBS_TARGET\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_REFERENCE_OBS_TARGET="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --reference_obs_target. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--reference_obs_target=*)
|
|
[ -n "$VIASH_PAR_REFERENCE_OBS_TARGET" ] && ViashError Bad arguments for option \'--reference_obs_target=*\': \'$VIASH_PAR_REFERENCE_OBS_TARGET\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_REFERENCE_OBS_TARGET=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--output)
|
|
[ -n "$VIASH_PAR_OUTPUT" ] && ViashError Bad arguments for option \'--output\': \'$VIASH_PAR_OUTPUT\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_OUTPUT="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --output. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--output=*)
|
|
[ -n "$VIASH_PAR_OUTPUT" ] && ViashError Bad arguments for option \'--output=*\': \'$VIASH_PAR_OUTPUT\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_OUTPUT=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--output_compression)
|
|
[ -n "$VIASH_PAR_OUTPUT_COMPRESSION" ] && ViashError Bad arguments for option \'--output_compression\': \'$VIASH_PAR_OUTPUT_COMPRESSION\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_OUTPUT_COMPRESSION="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --output_compression. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--output_compression=*)
|
|
[ -n "$VIASH_PAR_OUTPUT_COMPRESSION" ] && ViashError Bad arguments for option \'--output_compression=*\': \'$VIASH_PAR_OUTPUT_COMPRESSION\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_OUTPUT_COMPRESSION=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--output_obs_predictions)
|
|
[ -n "$VIASH_PAR_OUTPUT_OBS_PREDICTIONS" ] && ViashError Bad arguments for option \'--output_obs_predictions\': \'$VIASH_PAR_OUTPUT_OBS_PREDICTIONS\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_OUTPUT_OBS_PREDICTIONS="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --output_obs_predictions. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--output_obs_predictions=*)
|
|
[ -n "$VIASH_PAR_OUTPUT_OBS_PREDICTIONS" ] && ViashError Bad arguments for option \'--output_obs_predictions=*\': \'$VIASH_PAR_OUTPUT_OBS_PREDICTIONS\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_OUTPUT_OBS_PREDICTIONS=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--output_obs_probability)
|
|
[ -n "$VIASH_PAR_OUTPUT_OBS_PROBABILITY" ] && ViashError Bad arguments for option \'--output_obs_probability\': \'$VIASH_PAR_OUTPUT_OBS_PROBABILITY\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_OUTPUT_OBS_PROBABILITY="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --output_obs_probability. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--output_obs_probability=*)
|
|
[ -n "$VIASH_PAR_OUTPUT_OBS_PROBABILITY" ] && ViashError Bad arguments for option \'--output_obs_probability=*\': \'$VIASH_PAR_OUTPUT_OBS_PROBABILITY\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_OUTPUT_OBS_PROBABILITY=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--model)
|
|
[ -n "$VIASH_PAR_MODEL" ] && ViashError Bad arguments for option \'--model\': \'$VIASH_PAR_MODEL\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_MODEL="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --model. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--model=*)
|
|
[ -n "$VIASH_PAR_MODEL" ] && ViashError Bad arguments for option \'--model=*\': \'$VIASH_PAR_MODEL\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_MODEL=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
--max_iter)
|
|
[ -n "$VIASH_PAR_MAX_ITER" ] && ViashError Bad arguments for option \'--max_iter\': \'$VIASH_PAR_MAX_ITER\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_MAX_ITER="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to --max_iter. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
--max_iter=*)
|
|
[ -n "$VIASH_PAR_MAX_ITER" ] && ViashError Bad arguments for option \'--max_iter=*\': \'$VIASH_PAR_MAX_ITER\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_PAR_MAX_ITER=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
---engine)
|
|
VIASH_ENGINE_ID="$2"
|
|
shift 2
|
|
;;
|
|
---engine=*)
|
|
VIASH_ENGINE_ID="$(ViashRemoveFlags "$1")"
|
|
shift 1
|
|
;;
|
|
---setup)
|
|
VIASH_MODE='setup'
|
|
VIASH_SETUP_STRATEGY="$2"
|
|
shift 2
|
|
;;
|
|
---setup=*)
|
|
VIASH_MODE='setup'
|
|
VIASH_SETUP_STRATEGY="$(ViashRemoveFlags "$1")"
|
|
shift 1
|
|
;;
|
|
---dockerfile)
|
|
VIASH_MODE='dockerfile'
|
|
shift 1
|
|
;;
|
|
---docker_run_args)
|
|
VIASH_DOCKER_RUN_ARGS+=("$2")
|
|
shift 2
|
|
;;
|
|
---docker_run_args=*)
|
|
VIASH_DOCKER_RUN_ARGS+=("$(ViashRemoveFlags "$1")")
|
|
shift 1
|
|
;;
|
|
---docker_image_id)
|
|
VIASH_MODE='docker_image_id'
|
|
shift 1
|
|
;;
|
|
---debug)
|
|
VIASH_MODE='debug'
|
|
shift 1
|
|
;;
|
|
---cpus)
|
|
[ -n "$VIASH_META_CPUS" ] && ViashError Bad arguments for option \'---cpus\': \'$VIASH_META_CPUS\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_META_CPUS="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to ---cpus. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
---cpus=*)
|
|
[ -n "$VIASH_META_CPUS" ] && ViashError Bad arguments for option \'---cpus=*\': \'$VIASH_META_CPUS\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_META_CPUS=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
---memory)
|
|
[ -n "$VIASH_META_MEMORY" ] && ViashError Bad arguments for option \'---memory\': \'$VIASH_META_MEMORY\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_META_MEMORY="$2"
|
|
[ $# -lt 2 ] && ViashError Not enough arguments passed to ---memory. Use "--help" to get more information on the parameters. && exit 1
|
|
shift 2
|
|
;;
|
|
---memory=*)
|
|
[ -n "$VIASH_META_MEMORY" ] && ViashError Bad arguments for option \'---memory=*\': \'$VIASH_META_MEMORY\' \& \'$2\' - you should provide exactly one argument for this option. && exit 1
|
|
VIASH_META_MEMORY=$(ViashRemoveFlags "$1")
|
|
shift 1
|
|
;;
|
|
*) # positional arg or unknown option
|
|
# since the positional args will be eval'd, can we always quote, instead of using ViashQuote
|
|
VIASH_POSITIONAL_ARGS="$VIASH_POSITIONAL_ARGS '$1'"
|
|
[[ $1 == -* ]] && ViashWarning $1 looks like a parameter but is not a defined parameter and will instead be treated as a positional argument. Use "--help" to get more information on the parameters.
|
|
shift # past argument
|
|
;;
|
|
esac
|
|
done
|
|
|
|
# parse positional parameters
|
|
eval set -- $VIASH_POSITIONAL_ARGS
|
|
|
|
|
|
if [ "$VIASH_ENGINE_ID" == "native" ] ; then
|
|
VIASH_ENGINE_TYPE='native'
|
|
elif [ "$VIASH_ENGINE_ID" == "docker" ] ; then
|
|
VIASH_ENGINE_TYPE='docker'
|
|
else
|
|
ViashError "Engine '$VIASH_ENGINE_ID' is not recognized. Options are: docker, native."
|
|
exit 1
|
|
fi
|
|
|
|
if [[ "$VIASH_ENGINE_TYPE" == "docker" ]]; then
|
|
# check if docker is installed properly
|
|
ViashDockerInstallationCheck
|
|
|
|
# determine docker image id
|
|
if [[ "$VIASH_ENGINE_ID" == 'docker' ]]; then
|
|
VIASH_DOCKER_IMAGE_ID='images.viash-hub.com/vsh/openpipeline/annotate/onclass:fix-integration-tests'
|
|
fi
|
|
|
|
# print dockerfile
|
|
if [ "$VIASH_MODE" == "dockerfile" ]; then
|
|
ViashDockerfile "$VIASH_ENGINE_ID"
|
|
exit 0
|
|
|
|
elif [ "$VIASH_MODE" == "docker_image_id" ]; then
|
|
echo "$VIASH_DOCKER_IMAGE_ID"
|
|
exit 0
|
|
|
|
# enter docker container
|
|
elif [[ "$VIASH_MODE" == "debug" ]]; then
|
|
VIASH_CMD="docker run --entrypoint=bash ${VIASH_DOCKER_RUN_ARGS[@]} -v '$(pwd)':/pwd --workdir /pwd -t $VIASH_DOCKER_IMAGE_ID"
|
|
ViashNotice "+ $VIASH_CMD"
|
|
eval $VIASH_CMD
|
|
exit
|
|
|
|
# build docker image
|
|
elif [ "$VIASH_MODE" == "setup" ]; then
|
|
ViashDockerSetup "$VIASH_DOCKER_IMAGE_ID" "$VIASH_SETUP_STRATEGY"
|
|
ViashDockerCheckCommands "$VIASH_DOCKER_IMAGE_ID" 'bash'
|
|
exit 0
|
|
fi
|
|
|
|
# check if docker image exists
|
|
ViashDockerSetup "$VIASH_DOCKER_IMAGE_ID" ifneedbepullelsecachedbuild
|
|
ViashDockerCheckCommands "$VIASH_DOCKER_IMAGE_ID" 'bash'
|
|
fi
|
|
|
|
# setting computational defaults
|
|
|
|
# helper function for parsing memory strings
|
|
function ViashMemoryAsBytes {
|
|
local memory=`echo "$1" | tr '[:upper:]' '[:lower:]' | tr -d '[:space:]'`
|
|
local memory_regex='^([0-9]+)([kmgtp]i?b?|b)$'
|
|
if [[ $memory =~ $memory_regex ]]; then
|
|
local number=${memory/[^0-9]*/}
|
|
local symbol=${memory/*[0-9]/}
|
|
|
|
case $symbol in
|
|
b) memory_b=$number ;;
|
|
kb|k) memory_b=$(( $number * 1000 )) ;;
|
|
mb|m) memory_b=$(( $number * 1000 * 1000 )) ;;
|
|
gb|g) memory_b=$(( $number * 1000 * 1000 * 1000 )) ;;
|
|
tb|t) memory_b=$(( $number * 1000 * 1000 * 1000 * 1000 )) ;;
|
|
pb|p) memory_b=$(( $number * 1000 * 1000 * 1000 * 1000 * 1000 )) ;;
|
|
kib|ki) memory_b=$(( $number * 1024 )) ;;
|
|
mib|mi) memory_b=$(( $number * 1024 * 1024 )) ;;
|
|
gib|gi) memory_b=$(( $number * 1024 * 1024 * 1024 )) ;;
|
|
tib|ti) memory_b=$(( $number * 1024 * 1024 * 1024 * 1024 )) ;;
|
|
pib|pi) memory_b=$(( $number * 1024 * 1024 * 1024 * 1024 * 1024 )) ;;
|
|
esac
|
|
echo "$memory_b"
|
|
fi
|
|
}
|
|
# compute memory in different units
|
|
if [ ! -z ${VIASH_META_MEMORY+x} ]; then
|
|
VIASH_META_MEMORY_B=`ViashMemoryAsBytes $VIASH_META_MEMORY`
|
|
# do not define other variables if memory_b is an empty string
|
|
if [ ! -z "$VIASH_META_MEMORY_B" ]; then
|
|
VIASH_META_MEMORY_KB=$(( ($VIASH_META_MEMORY_B+999) / 1000 ))
|
|
VIASH_META_MEMORY_MB=$(( ($VIASH_META_MEMORY_KB+999) / 1000 ))
|
|
VIASH_META_MEMORY_GB=$(( ($VIASH_META_MEMORY_MB+999) / 1000 ))
|
|
VIASH_META_MEMORY_TB=$(( ($VIASH_META_MEMORY_GB+999) / 1000 ))
|
|
VIASH_META_MEMORY_PB=$(( ($VIASH_META_MEMORY_TB+999) / 1000 ))
|
|
VIASH_META_MEMORY_KIB=$(( ($VIASH_META_MEMORY_B+1023) / 1024 ))
|
|
VIASH_META_MEMORY_MIB=$(( ($VIASH_META_MEMORY_KIB+1023) / 1024 ))
|
|
VIASH_META_MEMORY_GIB=$(( ($VIASH_META_MEMORY_MIB+1023) / 1024 ))
|
|
VIASH_META_MEMORY_TIB=$(( ($VIASH_META_MEMORY_GIB+1023) / 1024 ))
|
|
VIASH_META_MEMORY_PIB=$(( ($VIASH_META_MEMORY_TIB+1023) / 1024 ))
|
|
else
|
|
# unset memory if string is empty
|
|
unset $VIASH_META_MEMORY_B
|
|
fi
|
|
fi
|
|
# unset nproc if string is empty
|
|
if [ -z "$VIASH_META_CPUS" ]; then
|
|
unset $VIASH_META_CPUS
|
|
fi
|
|
|
|
|
|
# check whether required parameters exist
|
|
if [ -z ${VIASH_PAR_INPUT+x} ]; then
|
|
ViashError '--input' is a required argument. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
if [ -z ${VIASH_PAR_CL_NLP_EMB_FILE+x} ]; then
|
|
ViashError '--cl_nlp_emb_file' is a required argument. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
if [ -z ${VIASH_PAR_CL_ONTOLOGY_FILE+x} ]; then
|
|
ViashError '--cl_ontology_file' is a required argument. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
if [ -z ${VIASH_PAR_CL_OBO_FILE+x} ]; then
|
|
ViashError '--cl_obo_file' is a required argument. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
if [ -z ${VIASH_PAR_REFERENCE_OBS_TARGET+x} ]; then
|
|
ViashError '--reference_obs_target' is a required argument. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
if [ -z ${VIASH_META_NAME+x} ]; then
|
|
ViashError 'name' is a required argument. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
if [ -z ${VIASH_META_FUNCTIONALITY_NAME+x} ]; then
|
|
ViashError 'functionality_name' is a required argument. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
if [ -z ${VIASH_META_RESOURCES_DIR+x} ]; then
|
|
ViashError 'resources_dir' is a required argument. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
if [ -z ${VIASH_META_EXECUTABLE+x} ]; then
|
|
ViashError 'executable' is a required argument. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
if [ -z ${VIASH_META_CONFIG+x} ]; then
|
|
ViashError 'config' is a required argument. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
if [ -z ${VIASH_META_TEMP_DIR+x} ]; then
|
|
ViashError 'temp_dir' is a required argument. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
|
|
# filling in defaults
|
|
if [ -z ${VIASH_PAR_MODALITY+x} ]; then
|
|
VIASH_PAR_MODALITY="rna"
|
|
fi
|
|
if [ -z ${VIASH_PAR_OUTPUT_OBS_PREDICTIONS+x} ]; then
|
|
VIASH_PAR_OUTPUT_OBS_PREDICTIONS="onclass_pred"
|
|
fi
|
|
if [ -z ${VIASH_PAR_OUTPUT_OBS_PROBABILITY+x} ]; then
|
|
VIASH_PAR_OUTPUT_OBS_PROBABILITY="onclass_prob"
|
|
fi
|
|
if [ -z ${VIASH_PAR_MAX_ITER+x} ]; then
|
|
VIASH_PAR_MAX_ITER="30"
|
|
fi
|
|
|
|
# check whether required files exist
|
|
if [ ! -z "$VIASH_PAR_INPUT" ] && [ ! -e "$VIASH_PAR_INPUT" ]; then
|
|
ViashError "Input file '$VIASH_PAR_INPUT' does not exist."
|
|
exit 1
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_CL_NLP_EMB_FILE" ] && [ ! -e "$VIASH_PAR_CL_NLP_EMB_FILE" ]; then
|
|
ViashError "Input file '$VIASH_PAR_CL_NLP_EMB_FILE' does not exist."
|
|
exit 1
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_CL_ONTOLOGY_FILE" ] && [ ! -e "$VIASH_PAR_CL_ONTOLOGY_FILE" ]; then
|
|
ViashError "Input file '$VIASH_PAR_CL_ONTOLOGY_FILE' does not exist."
|
|
exit 1
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_CL_OBO_FILE" ] && [ ! -e "$VIASH_PAR_CL_OBO_FILE" ]; then
|
|
ViashError "Input file '$VIASH_PAR_CL_OBO_FILE' does not exist."
|
|
exit 1
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_REFERENCE" ] && [ ! -e "$VIASH_PAR_REFERENCE" ]; then
|
|
ViashError "Input file '$VIASH_PAR_REFERENCE' does not exist."
|
|
exit 1
|
|
fi
|
|
|
|
# check whether parameters values are of the right type
|
|
if [[ -n "$VIASH_PAR_MAX_ITER" ]]; then
|
|
if ! [[ "$VIASH_PAR_MAX_ITER" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError '--max_iter' has to be an integer. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_CPUS" ]]; then
|
|
if ! [[ "$VIASH_META_CPUS" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'cpus' has to be an integer. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_MEMORY_B" ]]; then
|
|
if ! [[ "$VIASH_META_MEMORY_B" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'memory_b' has to be a long. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_MEMORY_KB" ]]; then
|
|
if ! [[ "$VIASH_META_MEMORY_KB" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'memory_kb' has to be a long. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_MEMORY_MB" ]]; then
|
|
if ! [[ "$VIASH_META_MEMORY_MB" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'memory_mb' has to be a long. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_MEMORY_GB" ]]; then
|
|
if ! [[ "$VIASH_META_MEMORY_GB" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'memory_gb' has to be a long. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_MEMORY_TB" ]]; then
|
|
if ! [[ "$VIASH_META_MEMORY_TB" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'memory_tb' has to be a long. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_MEMORY_PB" ]]; then
|
|
if ! [[ "$VIASH_META_MEMORY_PB" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'memory_pb' has to be a long. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_MEMORY_KIB" ]]; then
|
|
if ! [[ "$VIASH_META_MEMORY_KIB" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'memory_kib' has to be a long. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_MEMORY_MIB" ]]; then
|
|
if ! [[ "$VIASH_META_MEMORY_MIB" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'memory_mib' has to be a long. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_MEMORY_GIB" ]]; then
|
|
if ! [[ "$VIASH_META_MEMORY_GIB" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'memory_gib' has to be a long. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_MEMORY_TIB" ]]; then
|
|
if ! [[ "$VIASH_META_MEMORY_TIB" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'memory_tib' has to be a long. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
if [[ -n "$VIASH_META_MEMORY_PIB" ]]; then
|
|
if ! [[ "$VIASH_META_MEMORY_PIB" =~ ^[-+]?[0-9]+$ ]]; then
|
|
ViashError 'memory_pib' has to be a long. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
fi
|
|
|
|
# check whether value is belongs to a set of choices
|
|
if [ ! -z "$VIASH_PAR_OUTPUT_COMPRESSION" ]; then
|
|
VIASH_PAR_OUTPUT_COMPRESSION_CHOICES=("gzip;lzf")
|
|
IFS=';'
|
|
set -f
|
|
if ! [[ ";${VIASH_PAR_OUTPUT_COMPRESSION_CHOICES[*]};" =~ ";$VIASH_PAR_OUTPUT_COMPRESSION;" ]]; then
|
|
ViashError '--output_compression' specified value of \'$VIASH_PAR_OUTPUT_COMPRESSION\' is not in the list of allowed values. Use "--help" to get more information on the parameters.
|
|
exit 1
|
|
fi
|
|
set +f
|
|
unset IFS
|
|
fi
|
|
|
|
# create parent directories of output files, if so desired
|
|
if [ ! -z "$VIASH_PAR_OUTPUT" ] && [ ! -d "$(dirname "$VIASH_PAR_OUTPUT")" ]; then
|
|
mkdir -p "$(dirname "$VIASH_PAR_OUTPUT")"
|
|
fi
|
|
|
|
if [ "$VIASH_ENGINE_ID" == "native" ] ; then
|
|
if [ "$VIASH_MODE" == "run" ]; then
|
|
VIASH_CMD="bash"
|
|
else
|
|
ViashError "Engine '$VIASH_ENGINE_ID' does not support mode '$VIASH_MODE'."
|
|
exit 1
|
|
fi
|
|
fi
|
|
|
|
if [[ "$VIASH_ENGINE_TYPE" == "docker" ]]; then
|
|
# detect volumes from file arguments
|
|
VIASH_CHOWN_VARS=()
|
|
if [ ! -z "$VIASH_PAR_INPUT" ]; then
|
|
VIASH_DIRECTORY_MOUNTS+=( "$(ViashDockerAutodetectMountArg "$VIASH_PAR_INPUT")" )
|
|
VIASH_PAR_INPUT=$(ViashDockerAutodetectMount "$VIASH_PAR_INPUT")
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_CL_NLP_EMB_FILE" ]; then
|
|
VIASH_DIRECTORY_MOUNTS+=( "$(ViashDockerAutodetectMountArg "$VIASH_PAR_CL_NLP_EMB_FILE")" )
|
|
VIASH_PAR_CL_NLP_EMB_FILE=$(ViashDockerAutodetectMount "$VIASH_PAR_CL_NLP_EMB_FILE")
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_CL_ONTOLOGY_FILE" ]; then
|
|
VIASH_DIRECTORY_MOUNTS+=( "$(ViashDockerAutodetectMountArg "$VIASH_PAR_CL_ONTOLOGY_FILE")" )
|
|
VIASH_PAR_CL_ONTOLOGY_FILE=$(ViashDockerAutodetectMount "$VIASH_PAR_CL_ONTOLOGY_FILE")
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_CL_OBO_FILE" ]; then
|
|
VIASH_DIRECTORY_MOUNTS+=( "$(ViashDockerAutodetectMountArg "$VIASH_PAR_CL_OBO_FILE")" )
|
|
VIASH_PAR_CL_OBO_FILE=$(ViashDockerAutodetectMount "$VIASH_PAR_CL_OBO_FILE")
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_REFERENCE" ]; then
|
|
VIASH_DIRECTORY_MOUNTS+=( "$(ViashDockerAutodetectMountArg "$VIASH_PAR_REFERENCE")" )
|
|
VIASH_PAR_REFERENCE=$(ViashDockerAutodetectMount "$VIASH_PAR_REFERENCE")
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_OUTPUT" ]; then
|
|
VIASH_DIRECTORY_MOUNTS+=( "$(ViashDockerAutodetectMountArg "$VIASH_PAR_OUTPUT")" )
|
|
VIASH_PAR_OUTPUT=$(ViashDockerAutodetectMount "$VIASH_PAR_OUTPUT")
|
|
VIASH_CHOWN_VARS+=( "$VIASH_PAR_OUTPUT" )
|
|
fi
|
|
if [ ! -z "$VIASH_META_RESOURCES_DIR" ]; then
|
|
VIASH_DIRECTORY_MOUNTS+=( "$(ViashDockerAutodetectMountArg "$VIASH_META_RESOURCES_DIR")" )
|
|
VIASH_META_RESOURCES_DIR=$(ViashDockerAutodetectMount "$VIASH_META_RESOURCES_DIR")
|
|
fi
|
|
if [ ! -z "$VIASH_META_EXECUTABLE" ]; then
|
|
VIASH_DIRECTORY_MOUNTS+=( "$(ViashDockerAutodetectMountArg "$VIASH_META_EXECUTABLE")" )
|
|
VIASH_META_EXECUTABLE=$(ViashDockerAutodetectMount "$VIASH_META_EXECUTABLE")
|
|
fi
|
|
if [ ! -z "$VIASH_META_CONFIG" ]; then
|
|
VIASH_DIRECTORY_MOUNTS+=( "$(ViashDockerAutodetectMountArg "$VIASH_META_CONFIG")" )
|
|
VIASH_META_CONFIG=$(ViashDockerAutodetectMount "$VIASH_META_CONFIG")
|
|
fi
|
|
if [ ! -z "$VIASH_META_TEMP_DIR" ]; then
|
|
VIASH_DIRECTORY_MOUNTS+=( "$(ViashDockerAutodetectMountArg "$VIASH_META_TEMP_DIR")" )
|
|
VIASH_META_TEMP_DIR=$(ViashDockerAutodetectMount "$VIASH_META_TEMP_DIR")
|
|
fi
|
|
|
|
# get unique mounts
|
|
VIASH_UNIQUE_MOUNTS=($(for val in "${VIASH_DIRECTORY_MOUNTS[@]}"; do echo "$val"; done | sort -u))
|
|
fi
|
|
|
|
if [[ "$VIASH_ENGINE_TYPE" == "docker" ]]; then
|
|
# change file ownership
|
|
function ViashPerformChown {
|
|
if (( ${#VIASH_CHOWN_VARS[@]} )); then
|
|
set +e
|
|
VIASH_CMD="docker run --entrypoint=bash --rm ${VIASH_UNIQUE_MOUNTS[@]} $VIASH_DOCKER_IMAGE_ID -c 'chown $(id -u):$(id -g) --silent --recursive ${VIASH_CHOWN_VARS[@]}'"
|
|
ViashDebug "+ $VIASH_CMD"
|
|
eval $VIASH_CMD
|
|
set -e
|
|
fi
|
|
}
|
|
trap ViashPerformChown EXIT
|
|
fi
|
|
|
|
if [[ "$VIASH_ENGINE_TYPE" == "docker" ]]; then
|
|
# helper function for filling in extra docker args
|
|
if [ ! -z "$VIASH_META_MEMORY_B" ]; then
|
|
VIASH_DOCKER_RUN_ARGS+=("--memory=${VIASH_META_MEMORY_B}")
|
|
fi
|
|
if [ ! -z "$VIASH_META_CPUS" ]; then
|
|
VIASH_DOCKER_RUN_ARGS+=("--cpus=${VIASH_META_CPUS}")
|
|
fi
|
|
fi
|
|
|
|
if [[ "$VIASH_ENGINE_TYPE" == "docker" ]]; then
|
|
VIASH_CMD="docker run --entrypoint=bash ${VIASH_DOCKER_RUN_ARGS[@]} ${VIASH_UNIQUE_MOUNTS[@]} $VIASH_DOCKER_IMAGE_ID"
|
|
fi
|
|
|
|
|
|
# set dependency paths
|
|
|
|
|
|
ViashDebug "Running command: $(echo $VIASH_CMD)"
|
|
cat << VIASHEOF | eval $VIASH_CMD
|
|
set -e
|
|
tempscript=\$(mktemp "$VIASH_META_TEMP_DIR/viash-run-onclass-XXXXXX").py
|
|
function clean_up {
|
|
rm "\$tempscript"
|
|
}
|
|
function interrupt {
|
|
echo -e "\nCTRL-C Pressed..."
|
|
exit 1
|
|
}
|
|
trap clean_up EXIT
|
|
trap interrupt INT SIGINT
|
|
cat > "\$tempscript" << 'VIASHMAIN'
|
|
import sys
|
|
import logging
|
|
import mudata as mu
|
|
import anndata as ad
|
|
import re
|
|
import numpy as np
|
|
from OnClass.OnClassModel import OnClassModel
|
|
import obonet
|
|
from typing import Dict, Tuple
|
|
from tqdm import tqdm
|
|
|
|
|
|
## VIASH START
|
|
# The following code has been auto-generated by Viash.
|
|
par = {
|
|
'input': $( if [ ! -z ${VIASH_PAR_INPUT+x} ]; then echo "r'${VIASH_PAR_INPUT//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'modality': $( if [ ! -z ${VIASH_PAR_MODALITY+x} ]; then echo "r'${VIASH_PAR_MODALITY//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'input_layer': $( if [ ! -z ${VIASH_PAR_INPUT_LAYER+x} ]; then echo "r'${VIASH_PAR_INPUT_LAYER//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'cl_nlp_emb_file': $( if [ ! -z ${VIASH_PAR_CL_NLP_EMB_FILE+x} ]; then echo "r'${VIASH_PAR_CL_NLP_EMB_FILE//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'cl_ontology_file': $( if [ ! -z ${VIASH_PAR_CL_ONTOLOGY_FILE+x} ]; then echo "r'${VIASH_PAR_CL_ONTOLOGY_FILE//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'cl_obo_file': $( if [ ! -z ${VIASH_PAR_CL_OBO_FILE+x} ]; then echo "r'${VIASH_PAR_CL_OBO_FILE//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'var_query_gene_names': $( if [ ! -z ${VIASH_PAR_VAR_QUERY_GENE_NAMES+x} ]; then echo "r'${VIASH_PAR_VAR_QUERY_GENE_NAMES//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'reference': $( if [ ! -z ${VIASH_PAR_REFERENCE+x} ]; then echo "r'${VIASH_PAR_REFERENCE//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'reference_layer': $( if [ ! -z ${VIASH_PAR_REFERENCE_LAYER+x} ]; then echo "r'${VIASH_PAR_REFERENCE_LAYER//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'reference_obs_target': $( if [ ! -z ${VIASH_PAR_REFERENCE_OBS_TARGET+x} ]; then echo "r'${VIASH_PAR_REFERENCE_OBS_TARGET//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'output': $( if [ ! -z ${VIASH_PAR_OUTPUT+x} ]; then echo "r'${VIASH_PAR_OUTPUT//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'output_compression': $( if [ ! -z ${VIASH_PAR_OUTPUT_COMPRESSION+x} ]; then echo "r'${VIASH_PAR_OUTPUT_COMPRESSION//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'output_obs_predictions': $( if [ ! -z ${VIASH_PAR_OUTPUT_OBS_PREDICTIONS+x} ]; then echo "r'${VIASH_PAR_OUTPUT_OBS_PREDICTIONS//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'output_obs_probability': $( if [ ! -z ${VIASH_PAR_OUTPUT_OBS_PROBABILITY+x} ]; then echo "r'${VIASH_PAR_OUTPUT_OBS_PROBABILITY//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'model': $( if [ ! -z ${VIASH_PAR_MODEL+x} ]; then echo "r'${VIASH_PAR_MODEL//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'max_iter': $( if [ ! -z ${VIASH_PAR_MAX_ITER+x} ]; then echo "int(r'${VIASH_PAR_MAX_ITER//\'/\'\"\'\"r\'}')"; else echo None; fi )
|
|
}
|
|
meta = {
|
|
'name': $( if [ ! -z ${VIASH_META_NAME+x} ]; then echo "r'${VIASH_META_NAME//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'functionality_name': $( if [ ! -z ${VIASH_META_FUNCTIONALITY_NAME+x} ]; then echo "r'${VIASH_META_FUNCTIONALITY_NAME//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'resources_dir': $( if [ ! -z ${VIASH_META_RESOURCES_DIR+x} ]; then echo "r'${VIASH_META_RESOURCES_DIR//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'executable': $( if [ ! -z ${VIASH_META_EXECUTABLE+x} ]; then echo "r'${VIASH_META_EXECUTABLE//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'config': $( if [ ! -z ${VIASH_META_CONFIG+x} ]; then echo "r'${VIASH_META_CONFIG//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'temp_dir': $( if [ ! -z ${VIASH_META_TEMP_DIR+x} ]; then echo "r'${VIASH_META_TEMP_DIR//\'/\'\"\'\"r\'}'"; else echo None; fi ),
|
|
'cpus': $( if [ ! -z ${VIASH_META_CPUS+x} ]; then echo "int(r'${VIASH_META_CPUS//\'/\'\"\'\"r\'}')"; else echo None; fi ),
|
|
'memory_b': $( if [ ! -z ${VIASH_META_MEMORY_B+x} ]; then echo "int(r'${VIASH_META_MEMORY_B//\'/\'\"\'\"r\'}')"; else echo None; fi ),
|
|
'memory_kb': $( if [ ! -z ${VIASH_META_MEMORY_KB+x} ]; then echo "int(r'${VIASH_META_MEMORY_KB//\'/\'\"\'\"r\'}')"; else echo None; fi ),
|
|
'memory_mb': $( if [ ! -z ${VIASH_META_MEMORY_MB+x} ]; then echo "int(r'${VIASH_META_MEMORY_MB//\'/\'\"\'\"r\'}')"; else echo None; fi ),
|
|
'memory_gb': $( if [ ! -z ${VIASH_META_MEMORY_GB+x} ]; then echo "int(r'${VIASH_META_MEMORY_GB//\'/\'\"\'\"r\'}')"; else echo None; fi ),
|
|
'memory_tb': $( if [ ! -z ${VIASH_META_MEMORY_TB+x} ]; then echo "int(r'${VIASH_META_MEMORY_TB//\'/\'\"\'\"r\'}')"; else echo None; fi ),
|
|
'memory_pb': $( if [ ! -z ${VIASH_META_MEMORY_PB+x} ]; then echo "int(r'${VIASH_META_MEMORY_PB//\'/\'\"\'\"r\'}')"; else echo None; fi ),
|
|
'memory_kib': $( if [ ! -z ${VIASH_META_MEMORY_KIB+x} ]; then echo "int(r'${VIASH_META_MEMORY_KIB//\'/\'\"\'\"r\'}')"; else echo None; fi ),
|
|
'memory_mib': $( if [ ! -z ${VIASH_META_MEMORY_MIB+x} ]; then echo "int(r'${VIASH_META_MEMORY_MIB//\'/\'\"\'\"r\'}')"; else echo None; fi ),
|
|
'memory_gib': $( if [ ! -z ${VIASH_META_MEMORY_GIB+x} ]; then echo "int(r'${VIASH_META_MEMORY_GIB//\'/\'\"\'\"r\'}')"; else echo None; fi ),
|
|
'memory_tib': $( if [ ! -z ${VIASH_META_MEMORY_TIB+x} ]; then echo "int(r'${VIASH_META_MEMORY_TIB//\'/\'\"\'\"r\'}')"; else echo None; fi ),
|
|
'memory_pib': $( if [ ! -z ${VIASH_META_MEMORY_PIB+x} ]; then echo "int(r'${VIASH_META_MEMORY_PIB//\'/\'\"\'\"r\'}')"; else echo None; fi )
|
|
}
|
|
dep = {
|
|
|
|
}
|
|
|
|
## VIASH END
|
|
|
|
sys.path.append(meta["resources_dir"])
|
|
# START TEMPORARY WORKAROUND setup_logger
|
|
def setup_logger():
|
|
logger = logging.getLogger()
|
|
logger.setLevel(logging.INFO)
|
|
console_handler = logging.StreamHandler(sys.stdout)
|
|
logFormatter = logging.Formatter("%(asctime)s %(levelname)-8s %(message)s")
|
|
console_handler.setFormatter(logFormatter)
|
|
logger.addHandler(console_handler)
|
|
|
|
return logger
|
|
# END TEMPORARY WORKAROUND setup_logger
|
|
|
|
logger = setup_logger()
|
|
|
|
def map_celltype_to_ontology_id(cl_obo_file: str) -> Tuple[Dict[str, str], Dict[str, str]]:
|
|
"""
|
|
Map cell type names to ontology IDs and vice versa.
|
|
|
|
Parameters
|
|
----------
|
|
cl_obo_file : str
|
|
Path to the cell ontology file.
|
|
|
|
Returns
|
|
-------
|
|
Tuple[Dict[str, str], Dict[str, str]]
|
|
A tuple of two dictionaries. The first dictionary maps cell ontology IDs to cell type names.
|
|
The second dictionary maps cell type names to cell ontology IDs.
|
|
"""
|
|
graph = obonet.read_obo(cl_obo_file)
|
|
cl_id_to_name = {id_: data.get("name") for id_, data in graph.nodes(data=True)}
|
|
cl_id_to_name = {k: v for k, v in cl_id_to_name.items() if v is not None}
|
|
name_to_cl_id = {v: k for k, v in cl_id_to_name.items()}
|
|
return cl_id_to_name, name_to_cl_id
|
|
|
|
def predict_input_data(model: OnClassModel,
|
|
input_matrix: np.array,
|
|
input_modality: ad.AnnData,
|
|
id_to_name: dict,
|
|
obs_prediction: str,
|
|
obs_probability: str) -> ad.AnnData:
|
|
"""
|
|
Predict cell types for input data and save results to Anndata obj.
|
|
|
|
Parameters
|
|
----------
|
|
model : OnClassModel
|
|
The OnClass model.
|
|
input_matrix : np.array
|
|
The input data matrix.
|
|
input_modality : ad.AnnData
|
|
The input data Anndata object.
|
|
id_to_name : dict
|
|
Dictionary mapping cell ontology IDs to cell type names.
|
|
obs_prediction : str
|
|
The obs key for the predicted cell type.
|
|
obs_probability : str
|
|
The obs key for the predicted cell type probability.
|
|
|
|
Returns
|
|
-------
|
|
ad.AnnData
|
|
The input data Anndata object with the predicted cell types saved in obs.
|
|
"""
|
|
corr_test_feature = model.ProcessTestFeature(
|
|
test_feature=input_matrix,
|
|
test_genes=input_modality.var_names,
|
|
log_transform=False,
|
|
)
|
|
onclass_pred = model.Predict(corr_test_feature, use_normalize=False, refine=True, unseen_ratio=-1.0)
|
|
pred_label = [model.i2co[ind] for ind in onclass_pred[2]]
|
|
pred_cell_type_label = [id_to_name[id] for id in pred_label]
|
|
|
|
input_modality.obs[obs_prediction] = pred_cell_type_label
|
|
input_modality.obs[obs_probability] = np.max(onclass_pred[1], axis=1) / onclass_pred[1].sum(1)
|
|
return input_modality
|
|
|
|
def set_var_index(adata, var_name):
|
|
adata.var.index = [re.sub("\\\\.[0-9]+\$", "", s) for s in adata.var[var_name]]
|
|
return adata
|
|
|
|
def main():
|
|
|
|
if (not par["model"] and not par["reference"]) or (par["model"] and par["reference"]):
|
|
raise ValueError("Make sure to provide either 'model' or 'reference', but not both.")
|
|
|
|
logger.info("Reading input data")
|
|
input_mudata = mu.read_h5mu(par["input"])
|
|
input_modality = input_mudata.mod[par["modality"]].copy()
|
|
|
|
# Set var names to the desired gene name format (gene synbol, ensembl id, etc.)
|
|
input_modality = set_var_index(input_modality, par["var_query_gene_names"]) if par["var_query_gene_names"] else input_modality
|
|
input_matrix = input_modality.layers[par["input_layer"]].toarray() if par["input_layer"] else input_modality.X.toarray()
|
|
|
|
id_to_name, name_to_id = map_celltype_to_ontology_id(par["cl_obo_file"])
|
|
|
|
|
|
if par["model"]:
|
|
logger.info("Predicting cell types using pre-trained model")
|
|
model = OnClassModel(cell_type_nlp_emb_file=par["cl_nlp_emb_file"],
|
|
cell_type_network_file=par["cl_ontology_file"])
|
|
|
|
model.BuildModel(use_pretrain=par["model"], ngene=None)
|
|
|
|
|
|
elif par["reference"]:
|
|
logger.info("Reading reference data")
|
|
model = OnClassModel(cell_type_nlp_emb_file=par["cl_nlp_emb_file"],
|
|
cell_type_network_file=par["cl_ontology_file"])
|
|
|
|
reference_mudata = mu.read_h5mu(par["reference"])
|
|
reference_modality = reference_mudata.mod[par["modality"]].copy()
|
|
|
|
reference_modality.var["gene_symbol"] = list(reference_modality.var.index)
|
|
reference_modality.var.index = [re.sub("\\\\.[0-9]+\$", "", s) for s in reference_modality.var["ensemblid"]]
|
|
|
|
logger.info("Detecting common vars based on ensembl ids")
|
|
common_ens_ids = list(set(reference_modality.var.index).intersection(set(input_modality.var.index)))
|
|
|
|
logger.info(" reference n_vars: %i", reference_modality.n_vars)
|
|
logger.info(" input n_vars: %i", input_modality.n_vars)
|
|
logger.info(" intersect n_vars: %i", len(common_ens_ids))
|
|
assert len(common_ens_ids) >= 100, "The intersection of genes is too small."
|
|
|
|
reference_matrix = reference_modality.layers[par["reference_layer"]].toarray() if par["reference_layer"] else reference_modality.X.toarray()
|
|
|
|
logger.info("Training a model from reference...")
|
|
labels = reference_modality.obs[par["reference_obs_target"]].tolist()
|
|
labels_cl = [name_to_id[label] for label in labels]
|
|
_ = model.EmbedCellTypes(labels_cl)
|
|
(
|
|
corr_train_feature,
|
|
_,
|
|
corr_train_genes,
|
|
_,
|
|
) = model.ProcessTrainFeature(
|
|
train_feature=reference_matrix,
|
|
train_label=labels_cl,
|
|
train_genes=reference_modality.var_names,
|
|
test_feature=input_matrix,
|
|
test_genes=input_modality.var_names,
|
|
log_transform=False,
|
|
)
|
|
model.BuildModel(ngene=len(corr_train_genes))
|
|
model.Train(corr_train_feature,
|
|
labels_cl,
|
|
max_iter=par["max_iter"])
|
|
|
|
|
|
logger.info(f"Predicting cell types")
|
|
input_modality = predict_input_data(model,
|
|
input_matrix,
|
|
input_modality,
|
|
id_to_name,
|
|
par["output_obs_predictions"],
|
|
par["output_obs_probability"])
|
|
logger.info("Writing output data")
|
|
input_mudata.mod[par["modality"]] = input_modality
|
|
input_mudata.write_h5mu(par["output"], compression=par["output_compression"])
|
|
|
|
if __name__ == "__main__":
|
|
main()
|
|
VIASHMAIN
|
|
python -B "\$tempscript" &
|
|
wait "\$!"
|
|
|
|
VIASHEOF
|
|
|
|
|
|
if [[ "$VIASH_ENGINE_TYPE" == "docker" ]]; then
|
|
# strip viash automount from file paths
|
|
|
|
if [ ! -z "$VIASH_PAR_INPUT" ]; then
|
|
VIASH_PAR_INPUT=$(ViashDockerStripAutomount "$VIASH_PAR_INPUT")
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_CL_NLP_EMB_FILE" ]; then
|
|
VIASH_PAR_CL_NLP_EMB_FILE=$(ViashDockerStripAutomount "$VIASH_PAR_CL_NLP_EMB_FILE")
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_CL_ONTOLOGY_FILE" ]; then
|
|
VIASH_PAR_CL_ONTOLOGY_FILE=$(ViashDockerStripAutomount "$VIASH_PAR_CL_ONTOLOGY_FILE")
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_CL_OBO_FILE" ]; then
|
|
VIASH_PAR_CL_OBO_FILE=$(ViashDockerStripAutomount "$VIASH_PAR_CL_OBO_FILE")
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_REFERENCE" ]; then
|
|
VIASH_PAR_REFERENCE=$(ViashDockerStripAutomount "$VIASH_PAR_REFERENCE")
|
|
fi
|
|
if [ ! -z "$VIASH_PAR_OUTPUT" ]; then
|
|
VIASH_PAR_OUTPUT=$(ViashDockerStripAutomount "$VIASH_PAR_OUTPUT")
|
|
fi
|
|
if [ ! -z "$VIASH_META_RESOURCES_DIR" ]; then
|
|
VIASH_META_RESOURCES_DIR=$(ViashDockerStripAutomount "$VIASH_META_RESOURCES_DIR")
|
|
fi
|
|
if [ ! -z "$VIASH_META_EXECUTABLE" ]; then
|
|
VIASH_META_EXECUTABLE=$(ViashDockerStripAutomount "$VIASH_META_EXECUTABLE")
|
|
fi
|
|
if [ ! -z "$VIASH_META_CONFIG" ]; then
|
|
VIASH_META_CONFIG=$(ViashDockerStripAutomount "$VIASH_META_CONFIG")
|
|
fi
|
|
if [ ! -z "$VIASH_META_TEMP_DIR" ]; then
|
|
VIASH_META_TEMP_DIR=$(ViashDockerStripAutomount "$VIASH_META_TEMP_DIR")
|
|
fi
|
|
fi
|
|
|
|
|
|
# check whether required files exist
|
|
if [ ! -z "$VIASH_PAR_OUTPUT" ] && [ ! -e "$VIASH_PAR_OUTPUT" ]; then
|
|
ViashError "Output file '$VIASH_PAR_OUTPUT' does not exist."
|
|
exit 1
|
|
fi
|
|
|
|
|
|
exit 0
|