def action_parameters(space, version=None, oformat='text'):
''' Returns a JSON with whole results info for a given space and version '''
if space is None:
return False, 'Empty space label'
from flame.parameters import Parameters
param = Parameters()
param.loadYaml(space, version, isSpace=True)
if oformat != 'text':
return True, param
else:
yaml = param.dumpYAML()
for line in yaml:
print (line)
return True, 'parameters listed'
def action_parameters(model, version=None, oformat='text'):
''' Returns an object with whole results info for a given model and version '''
if model is None:
return False, 'Empty model label'
from flame.parameters import Parameters
param = Parameters()
success, results = param.loadYaml(model, version)
if not success:
print(f'error obtaining parametes for model {model} : {results}')
return False, results
if oformat != 'text':
return True, param
else:
yaml = param.dumpYAML()
for line in yaml:
print(line)
return True, 'parameters listed'
class Search:
def __init__(self, space, version, output_format=None, label=None):
LOG.debug('Starting search...')
self.space = space
self.version = version
self.label = label
self.param = Parameters()
self.conveyor = Conveyor()
# identify the workflow type
self.conveyor.setOrigin('sapply')
# load modelID
path = utils.space_path(space, version)
meta = os.path.join(path,'space-meta.pkl')
try:
with open(meta, 'rb') as handle:
modelID = pickle.load(handle)
except:
LOG.critical(f'Unable to load modelID from {meta}. Aborting...')
sys.exit()
self.conveyor.addMeta('modelID', modelID)
LOG.debug (f'Loaded space with modelID: {modelID}')
# assign prediction (search) label
self.conveyor.addVal(label, 'prediction_label', 'prediction label',
'method', 'single',
'Label used to identify the prediction')
success, results = self.param.loadYaml(space, version, isSpace=True)
if not success:
LOG.critical(f'Unable to load space parameters. {results}. Aborting...')
sys.exit()
# add additional output formats included in the constructor
# this is requiered to add JSON format as output when the object is
# instantiated from a web service call, requiring this output
if output_format != None:
if output_format not in self.param.getVal('output_format'):
self.param.appVal('output_format',output_format)
return
def set_single_CPU(self) -> None:
''' Forces the use of a single CPU '''
LOG.debug('parameter "numCPUs" forced to be 1')
self.param.setVal('numCPUs',1)
def getVal (self, idict, ikey):
if not ikey in idict:
return None
return idict[ikey]
# def run(self, input_source, runtime_param=None, metric=None, numsel=None, cutoff=None):
def run(self, param_dict):
''' Executes a default predicton workflow '''
metric = None
numsel = None
cutoff = None
# path to endpoint
epd = utils.space_path(self.space, self.version)
if not os.path.isdir(epd):
LOG.error(f'Unable to find space {self.space}')
self.conveyor.setError(f'Unable to find space {self.space}, version {self.version}')
if self.getVal(param_dict,'smarts') is not None:
input_source = param_dict['smarts']
self.param.setVal('input_type', 'smarts')
elif self.getVal(param_dict,'infile') is not None:
input_source = param_dict['infile']
else:
LOG.error(f'Unable to find input_file')
self.conveyor.setError('wrong format in the runtime similarity parameters')
if 'runtime_param' in param_dict:
runtime_param = self.getVal(param_dict, 'runtime_param')
if runtime_param is not None:
LOG.info (f'runtime parameters: {str(runtime_param)}')
try:
with open(runtime_param, 'r') as pfile:
rtparam = yaml.safe_load(pfile)
try:
metric = rtparam['similarity_metric']
numsel = rtparam['similarity_cutoff_num']
cutoff = rtparam['similarity_cutoff_distance']
except:
LOG.error('wrong format in the runtime similarity parameters')
self.conveyor.setError('wrong format in the runtime similarity parameters')
except:
LOG.error('runtime similarity parameter file not found')
self.conveyor.setError('runtime similarity parameter file not found')
else:
try:
metric = param_dict['metric']
numsel = param_dict['numsel']
cutoff = param_dict['cutoff']
except:
LOG.error('wrong format in the runtime similarity parameters')
self.conveyor.setError('wrong format in the runtime similarity parameters')
md = self.param.getVal('computeMD_method')
if utils.isFingerprint(md) and len(md) > 1:
LOG.warning(f'When using fingerprints, only a single type of MD can be used to build spaces. Selecting {md[0]}')
self.conveyor.setWarning(f'When using fingerprints, only a single type of MD can be used to build spaces. Selecting {md[0]}')
self.param.setVal('computeMD_method',[md[0]])
if not self.conveyor.getError():
# uses the child classes within the 'space' folder,
# to allow customization of
# the processing applied to each space
modpath = utils.smodule_path(self.space, self.version)
idata_child = importlib.import_module(modpath+".idata_child")
sapply_child = importlib.import_module(modpath+".sapply_child")
odata_child = importlib.import_module(modpath+".odata_child")
# run idata object, in charge of generate space data from input
try:
idata = idata_child.IdataChild(self.param, self.conveyor, input_source)
except:
LOG.warning ('Idata child architecture mismatch, defaulting to Idata parent')
idata = Idata(self.param, self.conveyor, input_source)
idata.run()
LOG.debug(f'idata child {type(idata).__name__} completed `run()`')
if not self.conveyor.getError():
# make sure there is X data
if not self.conveyor.isKey('xmatrix'):
if not self.conveyor.isKey ('SMARTS'):
LOG.debug(f'Failed to compute MDs')
self.conveyor.setError(f'Failed to compute MDs')
if not self.conveyor.getError():
# run apply object, in charge of generate a prediction from idata
try:
sapply = sapply_child.SapplyChild(self.param, self.conveyor)
except:
LOG.warning ('Sapply child architecture mismatch, defaulting to Sapply parent')
sapply = Sapply(self.param, self.conveyor)
sapply.run(cutoff, numsel, metric)
LOG.debug(f'sapply child {type(sapply).__name__} completed `run()`')
# run odata object, in charge of formatting the prediction results
# note that if any of the above steps failed, an error has been inserted in the
# conveyor and odata will take case of showing an error message
try:
odata = odata_child.OdataChild(self.param, self.conveyor)
except:
LOG.warning ('Odata child architecture mismatch, defaulting to Odata parent')
odata = Odata(self.param, self.conveyor)
return odata.run()
class Predict:
def __init__(self, model, version=0, output_format=None, label=None):
LOG.debug('Starting predict...')
self.model = model
self.version = version
self.param = Parameters()
self.conveyor = Conveyor()
# identify the workflow type
self.conveyor.setOrigin('apply')
# load modelID
success, result = utils.getModelID(model, version, 'model')
if not success:
LOG.critical(f'{result}. Aborting...')
sys.exit()
self.conveyor.addMeta('modelID', result)
LOG.debug (f'Loaded model with modelID: {result}')
# assign prediction label
self.conveyor.addVal(label, 'prediction_label', 'prediction label',
'method', 'single',
'Label used to identify the prediction')
success, results = self.param.loadYaml(model, version)
if not success:
LOG.critical(f'Unable to load model parameters. {results}. Aborting...')
sys.exit()
# add additional output formats included in the constructor
# this is requiered to add JSON format as output when the object is
# instantiated from a web service call, requiring this output
if output_format != None:
if output_format not in self.param.getVal('output_format'):
self.param.appVal('output_format',output_format)
if 'ghost' in output_format:
self.param.setVal('output_similar', False)
return
def get_ensemble(self):
''' Returns a Boolean indicating if the model uses external input
sources and a list with these sources '''
return self.param.getEnsemble()
def set_single_CPU(self) -> None:
''' Forces the use of a single CPU '''
LOG.debug('parameter "numCPUs" forced to be 1')
self.param.setVal('numCPUs',1)
def run(self, input_source):
''' Executes a default predicton workflow '''
# path to endpoint
endpoint = utils.model_path(self.model, self.version)
# if not os.path.isdir(endpoint):
# self.conveyor.setError(f'Unable to find model {self.model}, version {self.version}')
# #LOG.error(f'Unable to find model {self.model}')
# if not self.conveyor.getError():
# uses the child classes within the 'model' folder,
# to allow customization of
# the processing applied to each model
modpath = utils.module_path(self.model, self.version)
idata_child = importlib.import_module(modpath+".idata_child")
apply_child = importlib.import_module(modpath+".apply_child")
odata_child = importlib.import_module(modpath+".odata_child")
# run idata object, in charge of generate model data from input
try:
idata = idata_child.IdataChild(self.param, self.conveyor, input_source)
except:
LOG.warning ('Idata child architecture mismatch, defaulting to Idata parent')
idata = Idata(self.param, self.conveyor, input_source)
idata.run()
LOG.debug(f'idata child {type(idata).__name__} completed `run()`')
if not self.conveyor.getError():
success, results = idata.preprocess_apply()
if not success:
self.conveyor.setError(results)
if not self.conveyor.getError():
# make sure there is X data
if not self.conveyor.isKey('xmatrix'):
LOG.debug(f'Failed to compute MDs')
self.conveyor.setError(f'Failed to compute MDs')
# for secret models avoid searching similar compounds
space_pkl = os.path.join(endpoint,'space.pkl')
if not os.path.isfile(space_pkl):
self.param.setVal('output_similar', False)
if not self.conveyor.getError():
if self.param.getVal('output_similar') is True:
from flame.sapply import Sapply
metric = self.param.getVal('similarity_metric')
numsel = self.param.getVal('similarity_cutoff_num')
cutoff = self.param.getVal('similarity_cutoff_distance')
# sapply = Sapply(self.param, self.conveyor)
sapply_child = importlib.import_module(modpath+".sapply_child")
# run apply object, in charge of generate a prediction from idata
try:
sapply = sapply_child.SapplyChild(self.param, self.conveyor)
except:
LOG.warning ('Sapply child architecture mismatch, defaulting to Sapply parent')
sapply = Sapply(self.param, self.conveyor)
sapply.run(cutoff, numsel, metric)
LOG.debug(f'sapply child {type(sapply).__name__} completed `run()`')
if not self.conveyor.getError():
# run apply object, in charge of generate a prediction from idata
try:
apply = apply_child.ApplyChild(self.param, self.conveyor)
except:
LOG.warning ('Apply child architecture mismatch, defaulting to Apply parent')
apply = Apply(self.param, self.conveyor)
apply.run()
LOG.debug(f'apply child {type(apply).__name__} completed `run()`')
# run odata object, in charge of formatting the prediction results
# note that if any of the above steps failed, an error has been inserted in the
# conveyor and odata will take case of showing an error message
try:
odata = odata_child.OdataChild(self.param, self.conveyor)
except:
LOG.warning ('Odata child architecture mismatch, defaulting to Odata parent')
odata = Odata(self.param, self.conveyor)
return odata.run()
class Build:
def __init__(self,
model,
param_file=None,
param_string=None,
output_format=None):
LOG.debug('Starting build...')
self.model = model
self.param = Parameters()
self.conveyor = Conveyor()
# load parameters
if param_file is not None:
# use the param_file to update existing parameters at the model
# directory and save changes to make them persistent
success, message = self.param.delta(model,
0,
param_file,
iformat='YAML')
elif param_string is not None:
success, message = self.param.delta(model,
0,
param_string,
iformat='JSONS')
else:
# load parameter file at the model directory
success, message = self.param.loadYaml(model, 0)
# being unable to load parameters is a critical error
if not success:
LOG.critical(
f'Unable to load model parameters. "{message}" Aborting...')
sys.exit(1)
# add additional output formats included in the constructor
# this is requiered to add JSON format as output when the object is
# instantiated from a web service call, requiring this output
if output_format is not None:
if output_format not in self.param.getVal('output_format'):
self.param.appVal('output_format', output_format)
def get_ensemble(self):
''' Returns a Boolean indicating if the model uses external input
sources and a list with these sources '''
return self.param.getEnsemble()
def set_single_CPU(self) -> None:
''' Forces the use of a single CPU '''
LOG.debug('parameter "numCPUs" forced to be 1')
self.param.setVal('numCPUs', 1)
def run(self, input_source):
''' Executes a default predicton workflow '''
# path to endpoint
epd = utils.model_path(self.model, 0)
if not os.path.isdir(epd):
self.conveyor.setError(f'Unable to find model {self.model}')
#LOG.error(f'Unable to find model {self.model}')
# import ichild classes
if not self.conveyor.getError():
# uses the child classes within the 'model' folder,
# to allow customization of the processing applied to each model
modpath = utils.module_path(self.model, 0)
idata_child = importlib.import_module(modpath + ".idata_child")
learn_child = importlib.import_module(modpath + ".learn_child")
odata_child = importlib.import_module(modpath + ".odata_child")
# run idata object, in charge of generate model data from input
try:
idata = idata_child.IdataChild(self.param, self.conveyor,
input_source)
except:
LOG.warning(
'Idata child architecture mismatch, defaulting to Idata parent'
)
idata = Idata(self.param, self.conveyor, input_source)
idata.run()
LOG.debug(f'idata child {type(idata).__name__} completed `run()`')
if not self.conveyor.getError():
# check there is a suitable X and Y
if not self.conveyor.isKey('xmatrix'):
self.conveyor.setError(f'Failed to compute MDs')
if not self.conveyor.isKey('ymatrix'):
self.conveyor.setError(
f'No activity data (Y) found in training series')
if not self.conveyor.getError():
# instantiate learn (build a model from idata) and run it
learn = learn_child.LearnChild(self.param, self.conveyor)
learn.run()
try:
learn = learn_child.LearnChild(self.param, self.conveyor)
except:
LOG.warning(
'Learn child architecture mismatch, defaulting to Learn parent'
)
learn = Learn(self.param, self.conveyor)
LOG.debug(f'learn child {type(learn).__name__} completed `run()`')
# run odata object, in charge of formatting the prediction results
# note that if any of the above steps failed, an error has been inserted in the
# conveyor and odata will take case of showing an error message
try:
odata = odata_child.OdataChild(self.param, self.conveyor)
except:
LOG.warning(
'Odata child architecture mismatch, defaulting to Odata parent'
)
odata = Odata(self.param, self.conveyor)
return odata.run()
class Documentation:
''' Class storing the information needed to documentate models
Fields are loaded from a YAML file (documentation.yaml)
...
Attributes
----------
fields : dict
fields in the documentation
version : int
documentation version
Methods
-------
load_parameters()
Accesses to param file to retrieve all
information needed to document the model.
load_results()
Accesses to build results to retrieve all
information needed to document the model.
assign_parameters()
Fill documentation values corresponding to
model parameter values
assign_results()
Assign result values to documentation fields
get_upf_template()
creates a spreedsheet QMRF-like
get_prediction_template()
Creates a reporting document for predictions
'''
def __init__(self, model, version=0, context='model'):
''' Load the fields from the documentation file'''
self.model = model
self.version = version
self.fields = None
self.parameters = Parameters()
self.conveyor = None
# obtain the path and the default name of the model documents
documentation_file_path = utils.model_path(self.model, self.version)
documentation_file_name = os.path.join(documentation_file_path,
'documentation.yaml')
# load the main class dictionary (p) from this yaml file
if not os.path.isfile(documentation_file_name):
raise Exception('Documentation file not found')
try:
with open(documentation_file_name, 'r') as documentation_file:
self.fields = yaml.safe_load(documentation_file)
except Exception as e:
# LOG.error(f'Error loading documentation file with exception: {e}')
raise e
success, message = self.parameters.loadYaml(model, version)
if not success:
print(
f'Parameters could not be loaded. {message}. Please make sure endpoint and version are correct'
)
return
# Remove this after acc
#self.load_parameters()
if context == 'model':
self.load_results()
self.assign_parameters()
self.assign_results()
self.autocomplete_documentation()
self.setVal('md5', self.idataHash())
def safe_copy(inputfile):
''' this function makes sure that the input file contains only printable chars
'''
def delta(self, model, version, doc, iformat='YAML', isSpace=False):
''' load a set of parameters from the configuration file present
at the model directory
also, inserts the keys present in the param_file provided,
assuming that it contains a YAML-compatible format, like the one
generated by manage
adds some parameters identifying the model and the
hash of the configuration file
'''
# input is a string, either in JSON or YAML format
# this is the typical input sent by
if iformat not in ['JSON', 'JSONS', 'YAML', 'YAMLS']:
return False, 'input format not recognized'
if iformat == 'JSONS':
try:
newp = json.loads(doc)
except Exception as e:
return False, str(e)
elif iformat == 'YAMLS':
try:
newp = yaml.load(doc)
except Exception as e:
return False, str(e)
# input is a file, either in YAML or JSON format
else:
try:
with open(doc, 'r') as pfile:
if iformat == 'YAML':
newp = yaml.safe_load(pfile)
elif iformat == 'JSON':
newp = json.load(pfile)
except Exception as e:
return False, str(e)
# update interna dict with keys in the input file (delta)
black_list = []
for key in newp:
if key not in black_list:
val = newp[key]
# YAML define null values as 'None, which are interpreted
# as strings
if val == 'None':
val = None
if isinstance(val, dict):
for inner_key in val:
inner_val = val[inner_key]
if inner_val == 'None':
inner_val = None
self.setInnerVal(key, inner_key, inner_val)
#print ('@delta: adding',key, inner_key, inner_val)
else:
self.setVal(key, val)
#print ('@delta: adding',key,val,type(val))
# dump internal dict to the parameters file
if isSpace:
parameters_file_path = utils.space_path(model, version)
else:
parameters_file_path = utils.model_path(model, version)
parameters_file_name = os.path.join(parameters_file_path,
'documentation.yaml')
try:
with open(parameters_file_name, 'w') as pfile:
yaml.dump(self.fields, pfile)
except Exception as e:
return False, 'unable to write parameters'
self.setVal('md5', self.idataHash())
return True, 'OK'
def load_results(self):
'''
Load results pickle with model information
'''
# obtain the path and the default name of the results file
results_file_path = utils.model_path(self.model, self.version)
results_file_name = os.path.join(results_file_path,
'model-results.pkl')
self.conveyor = Conveyor()
# load the main class dictionary (p) from this yaml file
if not os.path.isfile(results_file_name):
raise Exception('Results file not found')
try:
with open(results_file_name, "rb") as input_file:
self.conveyor.load(input_file)
except Exception as e:
# LOG.error(f'No valid results pickle found at:
# {results_file_name}')
raise e
def getVal(self, key):
''' Return the value of the key parameter or None if it is
not found in the parameters dictionary
'''
if not key in self.fields:
return None
if 'value' in self.fields[key]:
return self.fields[key]['value']
return None
def getDict(self, key):
''' Return the value of the key parameter or None if it ises.
not found in the parameters dictionary
'''
d = {}
if not key in self.fields:
return d
element = self.fields[key]['value']
if isinstance(element, dict):
# iterate keys and copy to the temp dictionary
# the key and the content of 'value'
for k, v in element.items():
if 'value' in v:
d[k] = v['value']
return d
def setVal(self, key, value):
''' Sets the parameter defined by key to the given value
'''
# for existing keys, replace the contents of 'value'
if key in self.fields:
if "value" in self.fields[key]:
if not isinstance(self.fields[key]['value'], dict):
self.fields[key]["value"] = value
# this should never happen, since value is never a dictionary
# else:
# for k in value.keys():
# self.fields[key][k] = value[k]
# this behaviour is deprecated, do not add new keys
# # for new keys, create a new element with 'value' key
# else:
# self.fields[key] = {'value': value}
def setInnerVal(self, okey, ikey, value):
''' Sets a parameter within an internal dictionary. The entry is defined
by a key of the outer dictionary (okey) and a second key in the inner
dicctionary (ikey). The paramenter will be set to the given value
This function test the existence of all the keys and dictionaries to
prevent crashes and returns without setting the value if any error is
found
'''
if not okey in self.fields:
return
if not "value" in self.fields[okey]:
return
odict = self.fields[okey]['value']
if not isinstance(odict, dict):
return
# now we are sure that odict is the right inner dictionary
if not ikey in odict:
return
# algorithm parameters not present in the template
if not isinstance(odict[ikey], dict):
odict['value'] = value
return
# keys present in the template
if "value" in odict[ikey]:
odict[ikey]["value"] = value
def appVal(self, key, value):
''' Appends value to the end of existing key list
'''
if not key in self.fields:
return
if "value" in self.fields[key]:
vt = self.fields[key]['value']
# if the key is already a list, append the new value at the end
if isinstance(vt, list):
self.fields[key]['value'].append(value)
# ... otherwyse, create a list with the previous content and the
# new value
else:
self.fields[key]['value'] = [vt, value]
def dumpJSON(self):
return json.dumps(self.fields, allow_nan=True)
def dumpYAML(self):
yaml_out = []
order = [
'ID', 'Version', 'Model_title', 'Model_description', 'Keywords',
'Contact', 'Institution', 'Date', 'Endpoint', 'Endpoint_units',
'Interpretation', 'Dependent_variable', 'Species',
'Limits_applicability', 'Experimental_protocol',
'Model_availability', 'Data_info', 'Algorithm', 'Software',
'Descriptors', 'Algorithm_settings', 'AD_method', 'AD_parameters',
'Goodness_of_fit_statistics', 'Internal_validation_1',
'Internal_validation_2', 'External_validation', 'Comments',
'Other_related_models', 'Date_of_QMRF', 'Date_of_QMRF_updates',
'QMRF_updates', 'References', 'QMRF_same_models',
'Mechanistic_basis', 'Mechanistic_references',
'Supporting_information', 'Comment_on_the_endpoint',
'Endpoint_data_quality_and_variability', 'Descriptor_selection'
]
for ik in order:
if ik in self.fields:
k = ik
v = self.fields[k]
ivalue = ''
idescr = ''
ioptio = ''
## newest parameter formats are extended and contain
## rich metainformation for each entry
if 'value' in v:
if not isinstance(v['value'], dict):
ivalue = v['value']
else:
# print header of dictionary
yaml_out.append(f'{k} :')
# iterate keys assuming existence of value and description
for intk in v['value']:
intv = v['value'][intk]
if not isinstance(intv, dict):
yaml_out.append(
f' {intk:27} : {str(intv):30}'
) #{iioptio} {iidescr}')
else:
#print(intk)
intv = v['value'][intk]
iivalue = ''
if "value" in intv:
iivalue = intv["value"]
# else:
# iivalue = intv
iidescr = ''
if "description" in intv and intv[
"description"] is not None:
iidescr = intv["description"]
iioptio = ''
if 'options' in intv:
toptio = intv['options']
if isinstance(toptio, list):
if toptio != [None]:
iioptio = f' {toptio}'
if isinstance(iivalue, float):
iivalue = f'{iivalue:f}'
elif iivalue is None:
iivalue = ''
yaml_out.append(
f' {intk:27} : {str(iivalue):30} #{iioptio} {iidescr}'
)
continue
if 'description' in v:
idescr = v['description']
if 'options' in v:
toptio = v['options']
if isinstance(toptio, list):
ioptio = f' {toptio}'
yaml_out.append(
f'{k:30} : {str(ivalue):30} #{ioptio} {idescr}')
return (yaml_out)
def dumpExcel(self, oname):
# openpyxl should be installed in the environment
# pip install openpyxl
from openpyxl import Workbook
from openpyxl.styles import Font, NamedStyle, Alignment
# from openpyxl.comments import Comment
wb = Workbook()
ws = wb.active
ws.title = f"Model {self.model} documentation"
alignment_style = Alignment(vertical='top', wrapText=True)
# Label Style
Label = NamedStyle(name="Label")
Label.font = Font(name='Calibri', size=11, bold=True)
Label.alignment = alignment_style
ws.column_dimensions['A'].width = 25.10
ws.column_dimensions['B'].width = 28.00
ws.column_dimensions['C'].width = 60.00
ws.column_dimensions['D'].width = 60.00
# sections of the document, specifying the document keys which will be listed
sections = [
('General model information', [
'ID', 'Version', 'Model_title', 'Model_description',
'Keywords', 'Contact', 'Institution', 'Date', 'Endpoint',
'Endpoint_units', 'Interpretation', 'Dependent_variable',
'Species', 'Limits_applicability', 'Experimental_protocol',
'Model_availability', 'Data_info'
]),
('Algorithm and software', [
'Algorithm', 'Software', 'Descriptors', 'Algorithm_settings',
'AD_method', 'AD_parameters', 'Goodness_of_fit_statistics',
'Internal_validation_1', 'Internal_validation_2',
'External_validation', 'Comments'
]),
('Other information', [
'Other_related_models', 'Date_of_QMRF', 'Date_of_QMRF_updates',
'QMRF_updates', 'References', 'QMRF_same_models',
'Mechanistic_basis', 'Mechanistic_references',
'Supporting_information', 'Comment_on_the_endpoint',
'Endpoint_data_quality_and_variability', 'Descriptor_selection'
])
]
#Save the position and name of the label for the first and last section
position = []
name = [sections[0][1][0], 'Other Comments']
count = 1
for isection in sections:
for ik in isection[1]:
label_k = ik.replace('_', ' ')
if label_k == 'Internal validation 2' or label_k == 'External validation':
ws[f"A{count}"] = label_k
ws[f'A{count}'].style = Label
else:
ws[f"B{count}"] = label_k
ws[f"B{count}"].style = Label
if ik in self.fields:
# set defaults for value
ivalue = ''
#v is the selected entry in the documentation dictionary
v = self.fields[ik]
## newest parameter formats are extended and contain
## rich metainformation for each entry
if 'value' in v:
ivalue = v['value']
if isinstance(ivalue, dict):
ws[f"A{count}"] = label_k
ws[f"A{count}"].style = Label
end = (count) + (len(ivalue) - 1)
for intk in ivalue:
label_ik = intk.replace('_', ' ')
# label_ik = intk.replace('_f', '').replace('_', ' ')
ws[f'B{count}'] = label_ik
ws[f'B{count}'].style = Label
intv = ivalue[intk]
if not isinstance(intv, dict):
iivalue = intv
if iivalue is None:
iivalue = " "
else:
intv = ivalue[intk]
iivalue = ''
if 'value' in intv:
iivalue = intv["value"]
if iivalue is None:
iivalue = ''
ws[f'D{count}'] = intv['description']
ws[f'D{count}'].alignment = alignment_style
ws[f'C{count}'] = f'{str(iivalue)}'
ws[f'C{count}'].font = Font(name='Calibri',
size=11,
color='3465a4')
ws[f'C{count}'].alignment = alignment_style
ws.merge_cells(f'A{count}:A{end}')
count += 1
else:
ws[f'D{count}'] = v['description']
ws[f'D{count}'].alignment = alignment_style
if label_k == 'Experimental protocol' or label_k == 'Comments':
position.append(count)
if ivalue is None:
ivalue = ''
ws[f'C{count}'] = f'{str(ivalue)}'
ws[f'C{count}'].font = Font(name='Calibri',
size=11,
color='3465a4')
ws[f'C{count}'].alignment = alignment_style
count += 1
itr = 0
for i in position:
if itr == 0:
ws[f'A{1}'] = name[itr]
ws[f"A{1}"].style = Label
ws.merge_cells(f'A{1}:A{i}')
else:
ws[f'A{i}'] = name[itr]
ws[f"A{i}"].style = Label
ws.merge_cells(f'A{i}:A{count-1}')
itr += 1
try:
wb.save(oname)
except:
return False, f'error saving document as {oname}'
return True, 'OK'
def dumpWORD(self, oname):
# python-docx should be installed in the environment
# pip install python-docx
from docx import Document
from docx.shared import Pt
from docx.shared import RGBColor
# most of the formatting is included in this template, where we
# redefined default styles for Normal, 'heading 1' and 'Table Grid'
#
# note that this template can be easily customized with a company
# or project logo
path = os.path.dirname(os.path.abspath(__file__))
path = os.path.join(path, 'children')
path = os.path.join(path, 'documentation_template.docx')
document = Document(path)
# define style for normal and heading 1
# normal_style = document.styles['Normal']
# normal_font = normal_style.font
# normal_font.name = 'Calibri'
# normal_font.size = Pt(10)
# heading_style = document.styles['heading 1']
# heading_font = heading_style.font
# heading_font.name = 'Calibri'
# heading_font.color.rgb = RGBColor(0x00, 0x00, 0x00)
# heading_font.size = Pt(12)
# withd of column 1 and 2
wcol1 = 1400000
wcol2 = 4200000
# withd of internal columns i and 2
wicol1 = 1200000
wicol2 = 2900000
# sections of the document, specifying the document keys which will be listed
sections = [
('General model information', [
'ID', 'Version', 'Model_title', 'Model_description',
'Keywords', 'Contact', 'Institution', 'Date', 'Endpoint',
'Endpoint_units', 'Interpretation', 'Dependent_variable',
'Species', 'Limits_applicability', 'Experimental_protocol',
'Model_availability', 'Data_info'
]),
('Algorithm and software', [
'Algorithm', 'Software', 'Descriptors', 'Algorithm_settings',
'AD_method', 'AD_parameters', 'Goodness_of_fit_statistics',
'Internal_validation_1', 'Internal_validation_2',
'External_validation', 'Comments'
]),
('Other information', [
'Other_related_models', 'Date_of_QMRF', 'Date_of_QMRF_updates',
'QMRF_updates', 'References', 'QMRF_same_models',
'Mechanistic_basis', 'Mechanistic_references',
'Supporting_information', 'Comment_on_the_endpoint',
'Endpoint_data_quality_and_variability', 'Descriptor_selection'
])
]
for isection in sections:
# heading with the section name
document.add_heading(isection[0], level=1)
# table with one row per key
table = document.add_table(rows=len(isection[1]), cols=2)
table.style = 'Table Grid'
table.autofit = False
count = 0
for ik in isection[1]:
# add a row and format two columns
row = table.rows[count]
row.cells[0].width = wcol1
row.cells[1].width = wcol2
label_k = ik.replace('_', ' ')
row.cells[0].text = f'{label_k}'
count = count + 1
# define value
if ik in self.fields:
# set defaults for value
ivalue = ''
# v is the selected entry in the documentation dictionary
v = self.fields[ik]
## newest parameter formats are extended and contain
## rich metainformation for each entry
if 'value' in v:
ivalue = v['value']
# if ivalue is a dictionary create a nested table and iterate
# to represent the keys within
if isinstance(ivalue, dict):
row.cells[0].text = f'{label_k}'
itable = row.cells[1].add_table(rows=len(ivalue),
cols=2)
itable.style = 'Table Grid'
itable.autofit = False
icount = 0
# iterate keys assuming existence of value and description
for intk in ivalue:
label_ik = intk.replace('_', ' ')
# label_ik = intk.replace('_f', '').replace('_', ' ')
irow = itable.rows[icount]
irow.cells[0].width = wicol1
irow.cells[1].width = wicol2
icount = icount + 1
intv = ivalue[intk]
if not isinstance(intv, dict):
iivalue = intv
else:
intv = ivalue[intk]
iivalue = ''
if "value" in intv:
iivalue = intv["value"]
if isinstance(iivalue, float):
iivalue = f'{iivalue:f}'
elif iivalue is None:
iivalue = ''
irow.cells[0].text = f'{label_ik}'
irow.cells[1].text = f'{str(iivalue)}'
# if the key is not a dictionary just insert the value inside
else:
if ivalue is None:
ivalue = ''
row.cells[1].text = f'{str(ivalue)}'
try:
document.save(oname)
except:
return False, f'error saving document as {oname}'
return True, 'OK'
def assign_parameters(self):
'''
Fill documentation values corresponding to model parameter values
'''
if not self.parameters:
raise ('Parameters were not loaded')
# self.fields['Algorithm']['subfields']['algorithm']['value'] = \
# self.parameters.getVal('model')
self.setInnerVal('Algorithm', 'algorithm',
self.parameters.getVal('model'))
if self.parameters.getVal('input_type') == 'molecule':
self.setInnerVal('Algorithm', 'descriptors',
self.parameters.getVal('computeMD_method'))
cv_method = f'{self.parameters.getVal("ModelValidationCV")} ({str(self.parameters.getVal("ModelValidationN"))})'
self.setInnerVal('Algorithm', 'cross-validation', cv_method)
features = self.parameters.getVal("feature_selection")
if features is not None:
features += f' ({self.parameters.getVal("feature_number")})'
self.setInnerVal('Descriptors', 'descriptors',
self.parameters.getVal('computeMD_method'))
self.setInnerVal('Descriptors', 'scaling',
self.parameters.getVal('modelAutoscaling'))
self.setInnerVal('Descriptors', 'selection_method', features)
elif self.parameters.getVal('input_type') == 'model_ensemble':
self.setInnerVal('Descriptors', 'descriptors', 'ensemble models')
if self.parameters.getVal('conformal'):
self.setInnerVal('AD_method', 'name', 'conformal prediction')
# self.setInnerVal('AD_parameters', 'confidence', f'{self.parameters.getVal("conformalConfidence")}')
conformal_settings_dict = {}
conformal_settings_dict['confidence'] = self.parameters.getVal(
"conformalConfidence")
conformal_settings = self.parameters.getVal('conformal_settings')
if conformal_settings is not None:
for key in conformal_settings:
conformal_settings_dict[key] = conformal_settings[key][
"value"]
self.fields['AD_parameters']['value'] = conformal_settings_dict
def assign_results(self):
'''
Assign result values to documentation fields
'''
# Accepted validation keys
# allowed = ['Conformal_accuracy', 'Conformal_mean_interval',
# 'Conformal_coverage', 'Conformal_accuracy',
# 'Q2', 'SDEP',
# 'SensitivityPred', 'SpecificityPred', 'MCCpred']
# gof_allowed = ['R2', 'SDEC', 'scoringR'
# 'Sensitivity', 'Specificity', 'MCC']
allowed = [
'Conformal_accuracy', 'Conformal_mean_interval',
'Conformal_coverage', 'Q2', 'SDEP', 'scoringP', 'Sensitivity',
'Specificity', 'MCC'
]
gof_allowed = [
'Conformal_accuracy_f', 'Conformal_mean_interval_f',
'Conformal_coverage_f', 'R2', 'SDEC', 'scoringR', 'Sensitivity_f',
'Specificity_f', 'MCC_f'
]
model_info = self.conveyor.getVal('model_build_info')
validation = self.conveyor.getVal('model_valid_info')
# print(model_info)
# The code below to filter the hyperparameters to be
# reported.
# Get parameter keys for the used estimator
#param_key = self.parameters.getVal('model') + '_parameters'
# Get parameter dictionary
#estimator_params = self.parameters.getDict(param_key)
self.fields['Algorithm_settings']['value'] = \
(self.conveyor.getVal('estimator_parameters'))
# print (self.conveyor.getVal('estimator_parameters'))
# Horrendous patch to solve backcompatibility problem
if 'subfields' in self.fields['Data_info']:
sub_label = 'subfields'
else:
sub_label = 'value'
self.fields['Data_info'][sub_label]['training_set_size']['value'] = \
model_info[0][2]
self.fields['Descriptors'][sub_label]['final_number']['value'] = \
model_info[1][2]
self.fields['Descriptors'][sub_label]['ratio']['value'] = \
'{:0.2f}'.format(model_info[1][2]/model_info[0][2])
internal_val = dict()
for stat in validation:
if stat[0] in allowed:
internal_val[stat[0]] = float("{0:.2f}".format(stat[2]))
if internal_val:
self.fields['Internal_validation_1']\
['value'] = internal_val
gof = dict()
for stat in validation:
if stat[0] in gof_allowed:
gof[stat[0]] = float("{0:.2f}".format(stat[2]))
if gof:
self.fields['Goodness_of_fit_statistics']\
['value'] = gof
def get_string(self, dictionary):
'''
Convert a dictionary (from documentation.yaml)
to string format for the model template
'''
text = ''
for key, val in dictionary.items():
text += f'{key} : {val["value"]}\n'
return text
def get_string2(self, dictionary):
'''
Convert a dictionary (from parameter file) to
string format for the model template
'''
text = ''
for key, val in dictionary.items():
try:
if isinstance(str(val), str):
text += f'{key} : {val}\n'
except:
continue
return text
def get_upf_template(self):
'''
This function creates a tabular model template based
on the QMRF document type
'''
template = pd.DataFrame()
template['ID'] = ['']
template['Version'] = ['']
template['Description'] = ['']
template['Contact'] = ['']
template['Institution'] = ['']
template['Date'] = ['']
template['Endpoint'] = ['']
template['Endpoint_units'] = ['']
template['Dependent_variable'] = ['']
template['Species'] = ['']
template['Limits_applicability'] = ['']
template['Experimental_protocol'] = ['']
template['Data_info'] = [
self.get_string(self.fields['Data_info']['subfields'])
]
template['Model_availability'] = [\
self.get_string(self.fields['Model_availability']
['subfields'])]
template['Algorithm'] = [
self.get_string(self.fields['Algorithm']['subfields'])
]
template['Software'] = [
self.get_string(self.fields['Software']['subfields'])
]
template['Descriptors'] = [
self.get_string(self.fields['Descriptors']['subfields'])
]
template['Algorithm_settings'] = [
self.get_string(self.fields['Algorithm_settings']['subfields'])
]
template['AD_method'] = [
self.get_string(self.fields['AD_method']['subfields'])
]
template['AD_parameters'] = [self.fields['AD_parameters']['value']]
template['Goodness_of_fit_statistics'] = [self.fields\
['Goodness_of_fit_statistics']['value']]
template['Internal_validation_1'] = [
self.fields['Internal_validation_1']['value']
]
template.to_csv('QMRF_template.tsv', sep='\t')
def get_upf_template2(self):
'''
This function creates a tabular model template based
on the QMRF document type
'''
fields = ['ID', 'Version', 'Contact', 'Institution',\
'Date', 'Endpoint', 'Endpoint_units', 'Dependent_variable', 'Species',\
'Limits_applicability', 'Experimental_protocol', 'Data_info',\
'Model_availability', 'Algorithm', 'Software', 'Descriptors',\
'Algorithm_settings', 'AD_method', 'AD_parameters',\
'Goodness_of_fit_statistics', 'Internal_validation_1' ]
template = pd.DataFrame(
columns=['Field', 'Parameter name', 'Parameter value'])
for field in fields:
try:
subfields = self.fields[field]['subfields']
except:
subfields = self.fields[field]['value']
if subfields is not None:
for index, subfield in enumerate(subfields):
field2 = ''
if index == 0:
field2 = field
else:
field2 = ""
value = str(subfields[subfield]['value'])
# None types are retrieved as str from yaml??
if value == "None":
value = ""
row = dict(zip(['Field', 'Parameter name', 'Parameter value'],\
[field2, subfield, value]))
template = template.append(row, ignore_index=True)
else:
value = str(self.fields[field]['value'])
if value == 'None':
value = ""
row = dict(zip(['Field', 'Parameter name', 'Parameter value'],\
[field, "", value]))
template = template.append(row, ignore_index=True)
template.to_csv('QMRF_template3.tsv', sep='\t', index=False)
def get_prediction_template(self):
'''
This function creates a tabular model template based
on the QMRF document type
'''
# obtain the path and the default name of the results file
results_file_path = utils.model_path(self.model, self.version)
results_file_name = os.path.join(results_file_path,
'prediction-results.pkl')
conveyor = Conveyor()
# load the main class dictionary (p) from this yaml file
if not os.path.isfile(results_file_name):
raise Exception('Results file not found')
try:
with open(results_file_name, "rb") as input_file:
conveyor.load(input_file)
except Exception as e:
# LOG.error(f'No valid results pickle found at: {results_file_name}')
raise e
# First get Name, Inchi and InChIkey
names = conveyor.getVal('obj_nam')
smiles = conveyor.getVal('SMILES')
inchi = [AllChem.MolToInchi(AllChem.MolFromSmiles(m)) for m in smiles]
inchikeys = [
AllChem.InchiToInchiKey(
AllChem.MolToInchi(AllChem.MolFromSmiles(m))) for m in smiles
]
predictions = []
applicability = []
if self.parameters['quantitative']['value']:
raise ('Prediction template for quantitative endpoints'
' not implemented yet')
if not self.parameters['conformal']['value']:
predictions = conveyor.getVal('values')
else:
c0 = np.asarray(conveyor.getVal('c0'))
c1 = np.asarray(conveyor.getVal('c1'))
predictions = []
for i, j in zip(c0, c1):
prediction = ''
if i == j:
prediction = 'out of AD'
applicability.append('out')
if i != j:
if i == True:
prediction = 'Inactive'
else:
prediction = 'Active'
applicability.append('in')
predictions.append(prediction)
# Now create the spreedsheats for prediction
# First write summary
summary = ("Study name\n" + "Endpoint\n" + "QMRF-ID\n" +
"(Target)Compounds\n" +
"Compounds[compounds]\tName\tInChiKey\n")
for name, inch in zip(names, inchikeys):
summary += f'\t{name}\t{inch}\n'
summary += ("\nFile\n" + "Author name\n" + "E-mail\n" + "Role\n" +
"Affiliation\n" + "Date\n")
with open('summary_document.tsv', 'w') as out:
out.write(summary)
# Now prediction details
# Pandas is used to ease the table creation.
reporting = pd.DataFrame()
reporting['InChI'] = inchi
reporting['CAS-RN'] = '-'
reporting['SMILES'] = smiles
reporting['prediction'] = predictions
reporting['Applicability_domain'] = applicability
reporting['reliability'] = '-'
reporting['Structural_analogue_1_CAS'] = '-'
reporting['Structural_analogue_1_smiles'] = '-'
reporting['Structural_analogue_1_source'] = '-'
reporting['Structural_analogue_1_experimental_value'] = '-'
reporting['Structural_analogue_2_CAS'] = '-'
reporting['Structural_analogue_2_smiles'] = '-'
reporting['Structural_analogue_2_source'] = '-'
reporting['Structural_analogue_2_experimental_value'] = '-'
reporting['Structural_analogue_3_CAS'] = '-'
reporting['Structural_analogue_3_smiles'] = '-'
reporting['Structural_analogue_3_source'] = '-'
reporting['Structural_analogue_3_experimental_value'] = '-'
reporting.to_csv('prediction_report.tsv', sep='\t', index=False)
def idataHash(self):
''' Create a md5 hash for a number of keys describing parameters
relevant for idata
This hash is compared between runs, to check wether idata must
recompute or not the MD
'''
# update with any new idata relevant parameter
keylist = [
'SDFile_name', 'SDFile_activity', 'SDFile_experimental',
'normalize_method', 'ionize_method', 'convert3D_method',
'computeMD_method', 'TSV_objnames', 'TSV_activity', 'input_type'
]
idata_params = []
for i in keylist:
idata_params.append(self.getVal(i))
# MD_settings is a dictionary, obtain and sort the keys+values
md_params = self.getDict('MD_settings')
md_list = []
for key in md_params:
# combine key + value in a single string
md_list.append(key + str(md_params[key]))
idata_params.append(md_list.sort())
# use picke as a buffered object, neccesary to generate the hexdigest
p = pickle.dumps(idata_params)
return hashlib.md5(p).hexdigest()
def empty_fields(self):
'''
This function checks which fields do not contain values
'''
emptyfields = []
for ik in self.fields:
v = self.fields[ik]
if 'value' in v:
ivalue = v['value']
if isinstance(ivalue, dict):
for intk in ivalue:
intv = ivalue[intk]
if not isinstance(intv, dict):
iivalue = intv
if iivalue is None or len(str(iivalue)) is 0:
emptyfields.append(intk)
else:
intv = ivalue[intk]
iivalue = ''
if intv["value"] is None or len(str(
intv["value"])) is 0:
emptyfields.append(intk)
else:
if ivalue is None or len(str(ivalue)) is 0:
emptyfields.append(ik)
return emptyfields
def get_mols(self):
return dict(
zip(self.conveyor.getVal("obj_nam"),
self.conveyor.getVal("SMILES")))
def autocomplete_documentation(self):
"""
Auto complete fields in model documentation
"""
#ID, Model identifier.
self.fields['ID']['value'] = utils.getModelID(self.model, self.version,
'model')[1]
#Version
self.fields['Version']['value'] = str(self.version)
#Date, Date of model development and Date of QMRF.
today = date.today().strftime("%B %d, %Y")
self.fields['Date']['value'] = today
self.fields['Date_of_QMRF']['value'] = today
#format, Format used(SDF,TSV)
if self.parameters.getVal('input_type') == 'data':
self.fields['Data_info']['value']['format']['value'] = 'TSV'
else:
self.fields['Data_info']['value']['format']['value'] = 'SDF'
#Algorithm, type: QSAR.
self.fields['Algorithm']['value']['type']['value'] = 'QSAR'
#Model, Main modelling program, version, description and license.
software = "Flame, 1.0rc3"
fieldsapplysoftware = ['model', 'descriptors', 'applicability_domain']
for field in fieldsapplysoftware:
if field == 'applicability_domain':
if self.parameters.getVal('conformal'):
self.fields['Software']['value'][field]['value'] = software
else:
self.fields['Software']['value'][field]['value'] = software
class Sbuild:
def __init__(self,
space,
param_file=None,
param_string=None,
output_format=None):
LOG.debug('Starting sbuild...')
self.space = space
self.param = Parameters()
self.conveyor = Conveyor()
# identify the workflow type
self.conveyor.setOrigin('slearn')
# generate a unique modelID
self.conveyor.addMeta('modelID', utils.id_generator())
LOG.debug(
f'Generated new space with modelID: {self.conveyor.getMeta("modelID")}'
)
# load parameters
if param_file is not None:
# use the param_file to update existing parameters at the space
# directory and save changes to make them persistent
success, message = self.param.delta(space,
0,
param_file,
iformat='YAML',
isSpace=True)
elif param_string is not None:
success, message = self.param.delta(space,
0,
param_string,
iformat='JSONS',
isSpace=True)
else:
# load parameter file at the space directory
success, message = self.param.loadYaml(space, 0, isSpace=True)
# being unable to load parameters is a critical error
if not success:
LOG.critical(
f'Unable to load space parameters. {message}. Aborting...')
sys.exit(1)
md = self.param.getVal('computeMD_method')
if utils.isFingerprint(md) and len(md) > 1:
LOG.warning(
f'When using fingerprints, only a single type of MD can be used to build spaces. Selecting {md[0]}'
)
self.conveyor.setWarning(
f'When using fingerprints, only a single type of MD can be used to build spaces. Selecting {md[0]}'
)
self.param.setVal('computeMD_method', [md[0]])
# add additional output formats included in the constructor
# this is requiered to add JSON format as output when the object is
# instantiated from a web service call, requiring this output
if output_format is not None:
if output_format not in self.param.getVal('output_format'):
self.param.appVal('output_format', output_format)
def set_single_CPU(self) -> None:
''' Forces the use of a single CPU '''
LOG.debug('parameter "numCPUs" forced to be 1')
self.param.setVal('numCPUs', 1)
def run(self, input_source):
''' Executes a default chemical space building workflow '''
# path to endpoint
epd = utils.space_path(self.space, 0)
if not os.path.isdir(epd):
self.conveyor.setError(f'Unable to find space {self.space}')
#LOG.error(f'Unable to find space {self.space}')
# import ichild classes
if not self.conveyor.getError():
# uses the child classes within the 'space' folder,
# to allow customization of the processing applied to each space
modpath = utils.smodule_path(self.space, 0)
idata_child = importlib.import_module(modpath + ".idata_child")
slearn_child = importlib.import_module(modpath + ".slearn_child")
odata_child = importlib.import_module(modpath + ".odata_child")
# run idata object, in charge of generate space data from input
try:
idata = idata_child.IdataChild(self.param, self.conveyor,
input_source)
except:
LOG.warning(
'Idata child architecture mismatch, defaulting to Idata parent'
)
idata = Idata(self.param, self.conveyor, input_source)
idata.run()
LOG.debug(f'idata child {type(idata).__name__} completed `run()`')
if not self.conveyor.getError():
success, results = idata.preprocess_create()
if not success:
self.conveyor.setError(results)
if not self.conveyor.getError():
# check there is a suitable X and Y
if not self.conveyor.isKey('xmatrix'):
self.conveyor.setError(f'Failed to compute MDs')
if not self.conveyor.getError():
# instantiate learn (build a space from idata) and run it
try:
slearn = slearn_child.SlearnChild(self.param, self.conveyor)
except:
LOG.warning(
'Slearn child architecture mismatch, defaulting to Learn parent'
)
slearn = Slearn(self.param, self.conveyor)
slearn.run()
LOG.debug(
f'slearn child {type(slearn).__name__} completed `run()`')
# run odata object, in charge of formatting the prediction results
# note that if any of the above steps failed, an error has been inserted in the
# conveyor and odata will take case of showing an error message
try:
odata = odata_child.OdataChild(self.param, self.conveyor)
except:
LOG.warning(
'Odata child architecture mismatch, defaulting to Odata parent'
)
odata = Odata(self.param, self.conveyor)
return odata.run()
def action_parameters(space, version=None, oformat='text'):
''' Returns a JSON with whole results info for a given space and version '''
if space is None:
return False, 'Empty space label'
from flame.parameters import Parameters
param = Parameters()
param.loadYaml(space, version, isSpace=True)
if oformat == 'JSON':
return True, param.dumpJSON()
else:
order = [
'input_type',
'quantitative',
'SDFile_activity',
'SDFile_name',
'SDFile_id',
'SDFile_experimental',
'normalize_method',
'ionize_method',
'convert3D_method',
'computeMD_method',
'model',
'modelAutoscaling',
'tune',
'conformal',
'conformalSignificance',
'ModelValidationCV',
'ModelValidationLC',
'ModelValidationN',
'ModelValidationP',
'output_format',
'output_md',
'TSV_activity',
'TSV_objnames',
'TSV_varnames',
'imbalance',
'feature_selection',
'feature_number',
'mol_batch',
'ensemble_models',
'ensemble_versions',
'numCPUs',
'verbose_error',
'modelingToolkit',
'endpoint',
'model_path',
#'md5',
'version'
]
order += [
'MD_settings', 'RF_parameters', 'RF_optimize', 'SVM_parameters',
'SVM_optimize', 'PLSDA_parameters', 'PLSDA_optimize',
'PLSR_parameters', 'PLSR_optimize', 'GNB_parameters'
]
for ik in order:
if ik in param.p:
k = ik
v = param.p[k]
ivalue = ''
idescr = ''
ioptio = ''
## newest parameter formats are extended and contain
## rich metainformation for each entry
if param.extended:
if 'value' in v:
if not isinstance(v['value'], dict):
ivalue = v['value']
else:
# print header of dictionaty
print(f'{k} :')
# iterate keys assuming existence of value and description
for intk in v['value']:
intv = v['value'][intk]
iivalue = ''
if "value" in intv:
iivalue = intv["value"]
iidescr = ''
if "description" in intv and intv[
"description"] is not None:
iidescr = intv["description"]
iioptio = ''
if 'options' in intv:
toptio = intv['options']
if isinstance(toptio, list):
if toptio != [None]:
iioptio = f' {toptio}'
if isinstance(iivalue, float):
iivalue = f'{iivalue:f}'
elif iivalue is None:
iivalue = ''
print(
f' {intk:27} : {str(iivalue):30} #{iioptio} {iidescr}'
)
continue
if 'description' in v:
idescr = v['description']
if 'options' in v:
toptio = v['options']
if isinstance(toptio, list):
ioptio = f' {toptio}'
### compatibility: old stile parameters
else:
if not isinstance(v, dict):
ivalue = v
else:
ivalue = '*dictionary*'
### end compatibility
if isinstance(ivalue, float):
ivalue = f'{ivalue:f}'
elif ivalue is None:
ivalue = ''
print(f'{k:30} : {str(ivalue):30} #{ioptio} {idescr}')
return True, 'parameters listed'
class Predict:
def __init__(self, model, version=0, output_format=None, label=None):
LOG.debug('Starting predict...')
self.model = model
self.version = version
self.param = Parameters()
self.conveyor = Conveyor()
self.conveyor.addVal(label, 'prediction_label', 'prediction label',
'method', 'single',
'Label used to identify the prediction')
if not self.param.loadYaml(model, version):
LOG.critical('Unable to load model parameters. Aborting...')
sys.exit()
# add additional output formats included in the constructor
# this is requiered to add JSON format as output when the object is
# instantiated from a web service call, requiring this output
if output_format != None:
if output_format not in self.param.getVal('output_format'):
self.param.appVal('output_format', output_format)
return
def get_ensemble(self):
''' Returns a Boolean indicating if the model uses external input
sources and a list with these sources '''
return self.param.getEnsemble()
def set_single_CPU(self) -> None:
''' Forces the use of a single CPU '''
LOG.debug('parameter "numCPUs" forced to be 1')
self.param.setVal('numCPUs', 1)
def run(self, input_source):
''' Executes a default predicton workflow '''
# path to endpoint
# path to endpoint
endpoint = utils.model_path(self.model, self.version)
if not os.path.isdir(endpoint):
self.conveyor.setError(
f'Unable to find model {self.model}, version {self.version}')
#LOG.error(f'Unable to find model {self.model}')
if not self.conveyor.getError():
# uses the child classes within the 'model' folder,
# to allow customization of
# the processing applied to each model
modpath = utils.module_path(self.model, self.version)
idata_child = importlib.import_module(modpath + ".idata_child")
apply_child = importlib.import_module(modpath + ".apply_child")
odata_child = importlib.import_module(modpath + ".odata_child")
# run idata object, in charge of generate model data from input
try:
idata = idata_child.IdataChild(self.param, self.conveyor,
input_source)
except:
LOG.warning(
'Idata child architecture mismatch, defaulting to Idata parent'
)
idata = Idata(self.param, self.conveyor, input_source)
idata.run()
LOG.debug(f'idata child {type(idata).__name__} completed `run()`')
if not self.conveyor.getError():
# make sure there is X data
if not self.conveyor.isKey('xmatrix'):
LOG.debug(f'Failed to compute MDs')
self.conveyor.setError(f'Failed to compute MDs')
if not self.conveyor.getError():
# run apply object, in charge of generate a prediction from idata
try:
apply = apply_child.ApplyChild(self.param, self.conveyor)
except:
LOG.warning(
'Apply child architecture mismatch, defaulting to Apply parent'
)
apply = Apply(self.param, self.conveyor)
apply.run()
LOG.debug(f'apply child {type(apply).__name__} completed `run()`')
# run odata object, in charge of formatting the prediction results
# note that if any of the above steps failed, an error has been inserted in the
# conveyor and odata will take case of showing an error message
try:
odata = odata_child.OdataChild(self.param, self.conveyor)
except:
LOG.warning(
'Odata child architecture mismatch, defaulting to Odata parent'
)
odata = Odata(self.param, self.conveyor)
return odata.run()
def action_refresh(model=None, version=None, GUI=False):
'''
Rebuild one or many models making use of existing parameter files and
locally stored training series.
'''
import flame.context as context
from flame.parameters import Parameters
# from flame.documentation import Documentation
import logging
if GUI:
token_file = os.path.join(tempfile.gettempdir(), 'refreshing_' + model)
# update token file with content 'working'
with open(token_file, 'w') as f:
f.write('Analyzing and sorting models...')
# list endpoints relevant for the arguments
if model is not None:
model_list = [model]
else:
model_root = pathlib.Path(utils.model_repository_path())
model_list = [x.stem for x in model_root.iterdir() if x.is_dir()]
# list versions relevant for the arguments
task_list = []
for imodel in model_list:
if version is not None:
task_list.append((imodel, version))
else:
model_root = pathlib.Path(utils.model_tree_path(imodel))
itask_list = [(imodel, utils.modeldir2ver(x.stem))
for x in model_root.iterdir() if x.is_dir()]
task_list += itask_list # use "+=" and not "append" to merge the new list with the old one
# analize task_list and add at the end ensemble models
# this is needed to have low models refreshed BEFORE refreshing the high models
# eliminating the need to refresh them recursively
LOG.info("Analyzing and sorting models...")
# make sure the lower models are in task_list and, if not, force the inclussion
for itask in task_list:
param = Parameters()
success, results = param.loadYaml(itask[0], itask[1])
if not success:
continue
if param.getVal('input_type') == 'model_ensemble':
ens_nams = param.getVal('ensemble_names')
ens_vers = param.getVal('ensemble_versions')
for i in range(len(ens_nams)):
iver = 0
inam = ens_nams[i]
if (i < len(ens_vers)):
iver = ens_vers[i]
if ((inam, iver)) not in task_list:
task_list.append((inam, iver))
# create separate lists for regular and ensemble models
# and add ensemble models at the end
# this needs to be carried out after the previos step because
# some of the lower level models could be an ensemble model
# itself
mol_list = []
ens_list = []
for itask in task_list:
param = Parameters()
success, results = param.loadYaml(itask[0], itask[1])
if not success:
mol_list.append(itask)
continue
if param.getVal('input_type') == 'model_ensemble':
ens_list.append(itask)
else:
mol_list.append(itask)
task_list = mol_list + ens_list
# show all models before stating
LOG.info(
"Starting model refreshing task for the following models and versions")
for itask in task_list:
LOG.info(f' model: {itask[0]} version: {itask[1]}')
LOG.info("This can take some time, please be patient...")
source_dir = os.path.dirname(os.path.abspath(__file__))
children_dir = os.path.join(source_dir, 'children')
master_parameters = os.path.join(children_dir, 'parameters.yaml')
master_documentation = os.path.join(children_dir, 'documentation.yaml')
# now send the build command for each task
for itask in task_list:
destinat_path = utils.model_path(itask[0], 0) # dev
if itask[1] != 0:
# move version to /dev for building
original_path = utils.model_path(itask[0], itask[1]) # veri
security_path = destinat_path + '_security' # dev_sec
shutil.move(destinat_path, security_path) # dev --> dev_sec
shutil.move(original_path, destinat_path) # veri --> dev
LOG.info(
f' refreshing model: {itask[0]} version: {itask[1]} ({task_list.index(itask)+1} of {len(task_list)})...'
)
if GUI:
with open(token_file, 'w') as f:
f.write(
f'model: {itask[0]} version: {itask[1]} ({task_list.index(itask)+1} of {len(task_list)})'
)
# dissable LOG output
logging.disable(logging.ERROR)
# update parameters
dump_parameters = os.path.join(destinat_path, 'parameters_dump.yaml')
success, param = action_parameters(itask[0], 0, oformat='bin')
if success:
param_yaml = param.dumpYAML()
with open(dump_parameters, 'w') as f:
for line in param_yaml:
f.write(line + '\n')
else:
LOG.info(
' ERROR: unable to merge parameters for model: {itask[0]} version: {itask[1]}'
)
dump_parameters = None
original_parameters = os.path.join(destinat_path, 'parameters.yaml')
shutil.copy(master_parameters, original_parameters)
#update documentation
dump_documentation = os.path.join(destinat_path,
'documentation_dump.yaml')
success, documentation = action_documentation(itask[0],
0,
doc_file=None,
oformat='bin')
original_documentation = os.path.join(destinat_path,
'documentation.yaml')
shutil.copy(master_documentation, original_documentation)
if success:
documentation_yaml = documentation.dumpYAML()
with open(dump_documentation, 'w') as f:
for line in documentation_yaml:
line = line.encode("ascii", "ignore")
line = line.decode("ascii", "ignore")
f.write(line + '\n')
s2, documentation = action_documentation(itask[0],
0,
doc_file=None,
oformat='bin')
s3, r3 = documentation.delta(itask[0], 0, dump_documentation)
else:
LOG.info(
' ERROR: unable to merge documentation for model: {itask[0]} version: {itask[1]}'
)
# rebuild the model
command_build = {
'endpoint': itask[0],
'infile': None,
'param_file': dump_parameters,
'incremental': False
}
success, results = context.build_cmd(command_build)
# enable LOG output
logging.disable(logging.NOTSET)
if itask[1] != 0:
shutil.move(destinat_path, original_path) # dev --> veri
shutil.move(security_path, destinat_path) # dev_sec --> dev
if not success:
LOG.error(results)
LOG.info("Model refreshing task finished")
if GUI:
# update token file with status 'ready'
with open(token_file, 'w') as f:
f.write('ready')
return True, 'OK'
class Documentation:
''' Class storing the information needed to documentate models
Fields are loaded from a YAML file (documentation.yaml)
...
Attributes
----------
fields : dict
fields in the documentation
version : int
documentation version
Methods
-------
load_parameters()
Accesses to param file to retrieve all
information needed to document the model.
load_results()
Accesses to build results to retrieve all
information needed to document the model.
assign_parameters()
Fill documentation values corresponding to
model parameter values
assign_results()
Assign result values to documentation fields
get_upf_template()
creates a spreedsheet QMRF-like
get_prediction_template()
Creates a reporting document for predictions
'''
def __init__(self, model, version=0, context='model'):
''' Load the fields from the documentation file'''
self.model = model
self.version = version
self.fields = None
self.parameters = Parameters()
self.conveyor = None
# obtain the path and the default name of the model documents
documentation_file_path = utils.model_path(self.model, self.version)
documentation_file_name = os.path.join(documentation_file_path,
'documentation.yaml')
# load the main class dictionary (p) from this yaml file
if not os.path.isfile(documentation_file_name):
raise Exception('Documentation file not found')
try:
with open(documentation_file_name, 'r') as documentation_file:
self.fields = yaml.safe_load(documentation_file)
except Exception as e:
# LOG.error(f'Error loading documentation file with exception: {e}')
raise e
success, message = self.parameters.loadYaml(model, 0)
if not success:
print(
'Parameters could not be loaded. Please assure endpoint is correct'
)
return
# Remove this after acc
#self.load_parameters()
if context == 'model':
self.load_results()
self.assign_parameters()
self.assign_results()
self.setVal('md5', self.idataHash())
def delta(self, model, version, doc, iformat='YAML', isSpace=False):
''' load a set of parameters from the configuration file present
at the model directory
also, inserts the keys present in the param_file provided,
assuming that it contains a YAML-compatible format, like the one
generated by manage
adds some parameters identifying the model and the
hash of the configuration file
'''
# if not self.loadYaml(model, version, isSpace):
# return False, 'file not found'
# parse parameter file assuning it will be in
# a YAML-compatible format
if iformat == 'JSONS':
try:
newp = json.loads(doc)
except Exception as e:
return False, e
else:
try:
with open(doc, 'r') as pfile:
if iformat == 'YAML':
newp = yaml.safe_load(pfile)
elif iformat == 'JSON':
newp = json.load(pfile)
except Exception as e:
return False, e
# update interna dict with keys in the input file (delta)
black_list = []
for key in newp:
if key not in black_list:
val = newp[key]
# YAML define null values as 'None, which are interpreted
# as strings
if val == 'None':
val = None
if isinstance(val, dict):
for inner_key in val:
inner_val = val[inner_key]
if inner_val == 'None':
inner_val = None
self.setInnerVal(key, inner_key, inner_val)
#print ('@delta: adding',key, inner_key, inner_val)
else:
self.setVal(key, val)
#print ('@delta: adding',key,val,type(val))
# dump internal dict to the parameters file
if isSpace:
parameters_file_path = utils.space_path(model, version)
else:
parameters_file_path = utils.model_path(model, version)
parameters_file_name = os.path.join(parameters_file_path,
'documentation.yaml')
try:
with open(parameters_file_name, 'w') as pfile:
yaml.dump(self.fields, pfile)
except Exception as e:
return False, 'unable to write parameters'
self.setVal('md5', self.idataHash())
return True, 'OK'
def load_results(self):
'''
Load results pickle with model information
'''
# obtain the path and the default name of the results file
results_file_path = utils.model_path(self.model, self.version)
results_file_name = os.path.join(results_file_path, 'results.pkl')
self.conveyor = Conveyor()
# load the main class dictionary (p) from this yaml file
if not os.path.isfile(results_file_name):
raise Exception('Results file not found')
try:
with open(results_file_name, "rb") as input_file:
self.conveyor.load(input_file)
except Exception as e:
# LOG.error(f'No valid results pickle found at:
# {results_file_name}')
raise e
def getVal(self, key):
''' Return the value of the key parameter or None if it is
not found in the parameters dictionary
'''
if not key in self.fields:
return None
if 'value' in self.fields[key]:
return self.fields[key]['value']
return None
def getDict(self, key):
''' Return the value of the key parameter or None if it ises.
not found in the parameters dictionary
'''
d = {}
if not key in self.fields:
return d
element = self.fields[key]['value']
if isinstance(element, dict):
# iterate keys and copy to the temp dictionary
# the key and the content of 'value'
for k, v in element.items():
if 'value' in v:
d[k] = v['value']
return d
def setVal(self, key, value):
''' Sets the parameter defined by key to the given value
'''
# for existing keys, replace the contents of 'value'
if key in self.fields:
if "value" in self.fields[key]:
if not isinstance(self.fields[key]['value'], dict):
self.fields[key]["value"] = value
else:
# print(key)
for k in value.keys():
self.fields[key][k] = value[k]
# for new keys, create a new element with 'value' key
else:
self.fields[key] = {'value': value}
def setInnerVal(self, okey, ikey, value):
''' Sets a parameter within an internal dictionary. The entry is defined
by a key of the outer dictionary (okey) and a second key in the inner
dicctionary (ikey). The paramenter will be set to the given value
This function test the existence of all the keys and dictionaries to
prevent crashes and returns without setting the value if any error is
found
'''
if not okey in self.fields:
return
if not "value" in self.fields[okey]:
return
odict = self.fields[okey]['value']
if not isinstance(odict, dict):
return
if not ikey in odict:
return
if not isinstance(odict[ikey], dict):
odict['value'] = value
return
if "value" in odict[ikey]:
odict[ikey]["value"] = value
else:
odict[ikey] = {'value': value}
def appVal(self, key, value):
''' Appends value to the end of existing key list
'''
if not key in self.fields:
return
if "value" in self.fields[key]:
vt = self.fields[key]['value']
# if the key is already a list, append the new value at the end
if isinstance(vt, list):
self.fields[key]['value'].append(value)
# ... otherwyse, create a list with the previous content and the
# new value
else:
self.fields[key]['value'] = [vt, value]
def dumpJSON(self):
return json.dumps(self.fields)
def assign_parameters(self):
'''
Fill documentation values corresponding to model parameter values
'''
if not self.parameters:
raise ('Parameters were not loaded')
# self.fields['Algorithm']['subfields']['algorithm']['value'] = \
# self.parameters.getVal('model')
self.setInnerVal('Algorithm', 'algorithm',
self.parameters.getVal('model'))
self.setInnerVal('Algorithm', 'descriptors',
self.parameters.getVal('computeMD_method'))
if self.parameters.getVal('conformal'):
self.setInnerVal('AD_method', 'name', 'conformal prediction')
self.setVal(
'AD_parameters', f'Conformal Significance '
f'{self.parameters.getVal("conformalSignificance")}')
def assign_results(self):
'''
Assign result values to documentation fields
'''
# Accepted validation keys
allowed = [
'Conformal_accuracy', 'Conformal_mean_interval', 'Sensitivity',
'Specificity', 'MCC', 'Conformal_coverage', 'Conformal_accuracy',
'Q2', 'SDEP', 'SensitivityPed', 'SpecificityPred',
'SpecificityPred', 'MCCpred', 'scoringR', 'R2', 'SDEC'
]
model_info = self.conveyor.getVal('model_build_info')
validation = self.conveyor.getVal('model_valid_info')
# The code below to filter the hyperparameters to be
# reported.
# Get parameter keys for the used estimator
#param_key = self.parameters.getVal('model') + '_parameters'
# Get parameter dictionary
#estimator_params = self.parameters.getDict(param_key)
self.fields['Algorithm_settings']['value'] = \
(self.conveyor.getVal('estimator_parameters'))
# Horrendous patch to solve backcompatibility problem
if 'subfields' in self.fields['Data_info']:
sub_label = 'subfields'
else:
sub_label = 'value'
self.fields['Data_info']\
[sub_label]['training_set_size']['value'] = \
model_info[0][2]
self.fields['Data_info']\
[sub_label]['training_set_size']['value'] = \
model_info[0][2]
self.fields['Descriptors']\
[sub_label]['final_number']['value'] = \
model_info[1][2]
self.fields['Descriptors']\
[sub_label]['ratio']['value'] = \
'{:0.2f}'.format(model_info[1][2]/model_info[0][2])
internal_val = dict()
for stat in validation:
if stat[0] in allowed:
internal_val[stat[0]] = float("{0:.2f}".format(stat[2]))
if internal_val:
self.fields['Internal_validation_1']\
['value'] = internal_val
def get_string(self, dictionary):
'''
Convert a dictionary (from documentation.yaml)
to string format for the model template
'''
text = ''
for key, val in dictionary.items():
text += f'{key} : {val["value"]}\n'
return text
def get_string2(self, dictionary):
'''
Convert a dictionary (from parameter file) to
string format for the model template
'''
text = ''
for key, val in dictionary.items():
try:
if isinstance(str(val), str):
text += f'{key} : {val}\n'
except:
continue
return text
def get_upf_template(self):
'''
This function creates a tabular model template based
on the QMRF document type
'''
template = pd.DataFrame()
template['ID'] = ['']
template['Version'] = ['']
template['Description'] = ['']
template['Contact'] = ['']
template['Institution'] = ['']
template['Date'] = ['']
template['Endpoint'] = ['']
template['Endpoint_units'] = ['']
template['Dependent_variable'] = ['']
template['Species'] = ['']
template['Limits_applicability'] = ['']
template['Experimental_protocol'] = ['']
template['Data_info'] = [
self.get_string(self.fields['Data_info']['subfields'])
]
template['Model_availability'] = [\
self.get_string(self.fields['Model_availability']
['subfields'])]
template['Algorithm'] = [
self.get_string(self.fields['Algorithm']['subfields'])
]
template['Software'] = [
self.get_string(self.fields['Software']['subfields'])
]
template['Descriptors'] = [
self.get_string(self.fields['Descriptors']['subfields'])
]
template['Algorithm_settings'] = [
self.get_string(self.fields['Algorithm_settings']['subfields'])
]
template['AD_method'] = [
self.get_string(self.fields['AD_method']['subfields'])
]
template['AD_parameters'] = [self.fields['AD_parameters']['value']]
template['Goodness_of_fit_statistics'] = [self.fields\
['Goodness_of_fit_statistics']['value']]
template['Internal_validation_1'] = [
self.fields['Internal_validation_1']['value']
]
template.to_csv('QMRF_template.tsv', sep='\t')
def get_upf_template2(self):
'''
This function creates a tabular model template based
on the QMRF document type
'''
fields = ['ID', 'Version', 'Contact', 'Institution',\
'Date', 'Endpoint', 'Endpoint_units', 'Dependent_variable', 'Species',\
'Limits_applicability', 'Experimental_protocol', 'Data_info',\
'Model_availability', 'Algorithm', 'Software', 'Descriptors',\
'Algorithm_settings', 'AD_method', 'AD_parameters',\
'Goodness_of_fit_statistics', 'Internal_validation_1' ]
template = pd.DataFrame(
columns=['Field', 'Parameter name', 'Parameter value'])
for field in fields:
try:
subfields = self.fields[field]['subfields']
except:
subfields = self.fields[field]['value']
if subfields is not None:
for index, subfield in enumerate(subfields):
field2 = ''
if index == 0:
field2 = field
else:
field2 = ""
value = str(subfields[subfield]['value'])
# None types are retrieved as str from yaml??
if value == "None":
value = ""
row = dict(zip(['Field', 'Parameter name', 'Parameter value'],\
[field2, subfield, value]))
template = template.append(row, ignore_index=True)
else:
value = str(self.fields[field]['value'])
if value == 'None':
value = ""
row = dict(zip(['Field', 'Parameter name', 'Parameter value'],\
[field, "", value]))
template = template.append(row, ignore_index=True)
template.to_csv('QMRF_template3.tsv', sep='\t', index=False)
def get_prediction_template(self):
'''
This function creates a tabular model template based
on the QMRF document type
'''
# obtain the path and the default name of the results file
results_file_path = utils.model_path(self.model, self.version)
results_file_name = os.path.join(results_file_path,
'prediction-results.pkl')
conveyor = Conveyor()
# load the main class dictionary (p) from this yaml file
if not os.path.isfile(results_file_name):
raise Exception('Results file not found')
try:
with open(results_file_name, "rb") as input_file:
conveyor.load(input_file)
except Exception as e:
# LOG.error(f'No valid results pickle found at: {results_file_name}')
raise e
# First get Name, Inchi and InChIkey
names = conveyor.getVal('obj_nam')
smiles = conveyor.getVal('SMILES')
inchi = [AllChem.MolToInchi(AllChem.MolFromSmiles(m)) for m in smiles]
inchikeys = [
AllChem.InchiToInchiKey(
AllChem.MolToInchi(AllChem.MolFromSmiles(m))) for m in smiles
]
predictions = []
applicability = []
if self.parameters['quantitative']['value']:
raise ('Prediction template for quantitative endpoints'
' not implemented yet')
if not self.parameters['conformal']['value']:
predictions = conveyor.getVal('values')
else:
c0 = np.asarray(conveyor.getVal('c0'))
c1 = np.asarray(conveyor.getVal('c1'))
predictions = []
for i, j in zip(c0, c1):
prediction = ''
if i == j:
prediction = 'out of AD'
applicability.append('out')
if i != j:
if i == True:
prediction = 'Inactive'
else:
prediction = 'Active'
applicability.append('in')
predictions.append(prediction)
# Now create the spreedsheats for prediction
# First write summary
summary = ("Study name\n" + "Endpoint\n" + "QMRF-ID\n" +
"(Target)Compounds\n" +
"Compounds[compounds]\tName\tInChiKey\n")
for name, inch in zip(names, inchikeys):
summary += f'\t{name}\t{inch}\n'
summary += ("\nFile\n" + "Author name\n" + "E-mail\n" + "Role\n" +
"Affiliation\n" + "Date\n")
with open('summary_document.tsv', 'w') as out:
out.write(summary)
# Now prediction details
# Pandas is used to ease the table creation.
reporting = pd.DataFrame()
reporting['InChI'] = inchi
reporting['CAS-RN'] = '-'
reporting['SMILES'] = smiles
reporting['prediction'] = predictions
reporting['Applicability_domain'] = applicability
reporting['reliability'] = '-'
reporting['Structural_analogue_1_CAS'] = '-'
reporting['Structural_analogue_1_smiles'] = '-'
reporting['Structural_analogue_1_source'] = '-'
reporting['Structural_analogue_1_experimental_value'] = '-'
reporting['Structural_analogue_2_CAS'] = '-'
reporting['Structural_analogue_2_smiles'] = '-'
reporting['Structural_analogue_2_source'] = '-'
reporting['Structural_analogue_2_experimental_value'] = '-'
reporting['Structural_analogue_3_CAS'] = '-'
reporting['Structural_analogue_3_smiles'] = '-'
reporting['Structural_analogue_3_source'] = '-'
reporting['Structural_analogue_3_experimental_value'] = '-'
reporting.to_csv('prediction_report.tsv', sep='\t', index=False)
def idataHash(self):
''' Create a md5 hash for a number of keys describing parameters
relevant for idata
This hash is compared between runs, to check wether idata must
recompute or not the MD
'''
# update with any new idata relevant parameter
keylist = [
'SDFile_name', 'SDFile_activity', 'SDFile_experimental',
'normalize_method', 'ionize_method', 'convert3D_method',
'computeMD_method', 'TSV_varnames', 'TSV_objnames', 'TSV_activity',
'input_type'
]
idata_params = []
for i in keylist:
idata_params.append(self.getVal(i))
# MD_settings is a dictionary, obtain and sort the keys+values
md_params = self.getDict('MD_settings')
md_list = []
for key in md_params:
# combine key + value in a single string
md_list.append(key + str(md_params[key]))
idata_params.append(md_list.sort())
# use picke as a buffered object, neccesary to generate the hexdigest
p = pickle.dumps(idata_params)
return hashlib.md5(p).hexdigest()
def action_refresh(model=None, version=None):
'''
Rebuild one or many models making use of existing parameter files and
locally stored training series.
'''
import flame.context as context
from flame.parameters import Parameters
import logging
# list endpoints relevant for the arguments
if model is not None:
model_list = [model]
else:
model_root = pathlib.Path(utils.model_repository_path())
model_list = [x.stem for x in model_root.iterdir() if x.is_dir()]
# list versions relevant for the arguments
task_list = []
for imodel in model_list:
if version is not None:
task_list.append((imodel, version))
else:
model_root = pathlib.Path(utils.model_tree_path(imodel))
itask_list = [(imodel, utils.modeldir2ver(x.stem))
for x in model_root.iterdir() if x.is_dir()]
task_list += itask_list # use "+=" and not "append" to merge the new list with the old one
# analize task_list and add at the end ensemble models
# this is needed to have low models refreshed BEFORE refreshing the high models
# eliminating the need to refresh them recursively
LOG.info("Analyzing and sorting models...")
# make sure the lower models are in task_list and, if not, force the inclussion
for itask in task_list:
param = Parameters()
success, results = param.loadYaml(itask[0], itask[1])
if not success:
continue
if param.getVal('input_type') == 'model_ensemble':
ens_nams = param.getVal('ensemble_names')
ens_vers = param.getVal('ensemble_versions')
for i in range(len(ens_nams)):
iver = 0
inam = ens_nams[i]
if (i < len(ens_vers)):
iver = ens_vers[i]
if ((inam, iver)) not in task_list:
task_list.append((inam, iver))
# create separate lists for regular and ensemble models
# and add ensemble models at the end
# this needs to be carried out after the previos step because
# some of the lower level models could be an ensemble model
# itself
mol_list = []
ens_list = []
for itask in task_list:
param = Parameters()
success, results = param.loadYaml(itask[0], itask[1])
if not success:
mol_list.append(itask)
continue
if param.getVal('input_type') == 'model_ensemble':
ens_list.append(itask)
else:
mol_list.append(itask)
task_list = mol_list + ens_list
# show all models before stating
LOG.info(
"Starting model refreshing task for the following models and versions")
for itask in task_list:
LOG.info(f' model: {itask[0]} version: {itask[1]}')
LOG.info("This can take some time, please be patient...")
# now send the build command for each task
for itask in task_list:
if itask[1] != 0:
# move version to /dev for building
original_path = utils.model_path(itask[0], itask[1]) # veri
destinat_path = utils.model_path(itask[0], 0) # dev
security_path = destinat_path + '_security' # dev_sec
shutil.move(destinat_path, security_path) # dev --> dev_sec
shutil.move(original_path, destinat_path) # veri --> dev
LOG.info(
f' refreshing model: {itask[0]} version: {itask[1]} ({task_list.index(itask)+1} of {len(task_list)})...'
)
# dissable LOG output
logging.disable(logging.ERROR)
command_build = {
'endpoint': itask[0],
'infile': None,
'param_file': None,
'incremental': False
}
success, results = context.build_cmd(command_build)
# enable LOG output
logging.disable(logging.NOTSET)
if itask[1] != 0:
shutil.move(destinat_path, original_path) # dev --> veri
shutil.move(security_path, destinat_path) # dev_sec --> dev
if not success:
LOG.error(results)
LOG.info("Model refreshing task finished")
return True, 'OK'
class Search:
def __init__(self, space, version, output_format=None, label=None):
LOG.debug('Starting predict...')
self.space = space
self.version = version
self.label = label
self.param = Parameters()
self.conveyor = Conveyor()
self.conveyor.addVal(label, 'prediction_label', 'prediction label',
'method', 'single',
'Label used to identify the prediction')
if not self.param.loadYaml(space, version, isSpace=True):
LOG.critical('Unable to load space parameters. Aborting...')
sys.exit()
# add additional output formats included in the constructor
# this is requiered to add JSON format as output when the object is
# instantiated from a web service call, requiring this output
if output_format != None:
if output_format not in self.param.getVal('output_format'):
self.param.appVal('output_format', output_format)
return
def set_single_CPU(self) -> None:
''' Forces the use of a single CPU '''
LOG.debug('parameter "numCPUs" forced to be 1')
self.param.setVal('numCPUs', 1)
# def run(self, input_source, runtime_param=None, metric=None, numsel=None, cutoff=None):
def run(self, param_dict):
''' Executes a default predicton workflow '''
print('*********', param_dict)
metric = None
numsel = None
cutoff = None
# path to endpoint
epd = utils.space_path(self.space, self.version)
if not os.path.isdir(epd):
self.conveyor.setError(
f'Unable to find space {self.space}, version {self.version}')
#LOG.error(f'Unable to find space {self.space}')
if 'infile' in param_dict:
input_source = param_dict['infile']
else:
LOG.error(f'Unable to find input_file')
self.conveyor.setError(
'wrong format in the runtime similarity parameters')
if 'runtime_param' in param_dict:
runtime_param = param_dict['runtime_param']
if runtime_param is not None:
print(runtime_param)
try:
with open(runtime_param, 'r') as pfile:
rtparam = yaml.safe_load(pfile)
try:
metric = rtparam['similarity_metric']
numsel = rtparam['similarity_cutoff_num']
cutoff = rtparam['similarity_cutoff_distance']
except:
LOG.error(
'wrong format in the runtime similarity parameters'
)
self.conveyor.setError(
'wrong format in the runtime similarity parameters'
)
except:
LOG.error('runtime similarity parameter file not found')
self.conveyor.setError(
'runtime similarity parameter file not found')
else:
try:
metric = param_dict['metric']
numsel = param_dict['numsel']
cutoff = param_dict['cutoff']
except:
LOG.error('wrong format in the runtime similarity parameters')
self.conveyor.setError(
'wrong format in the runtime similarity parameters')
if not self.conveyor.getError():
# uses the child classes within the 'space' folder,
# to allow customization of
# the processing applied to each space
modpath = utils.smodule_path(self.space, self.version)
idata_child = importlib.import_module(modpath + ".idata_child")
sapply_child = importlib.import_module(modpath + ".sapply_child")
odata_child = importlib.import_module(modpath + ".odata_child")
# run idata object, in charge of generate space data from input
try:
idata = idata_child.IdataChild(self.param, self.conveyor,
input_source)
except:
LOG.warning(
'Idata child architecture mismatch, defaulting to Idata parent'
)
idata = Idata(self.param, self.conveyor, input_source)
idata.run()
LOG.debug(f'idata child {type(idata).__name__} completed `run()`')
if not self.conveyor.getError():
# make sure there is X data
if not self.conveyor.isKey('xmatrix'):
LOG.debug(f'Failed to compute MDs')
self.conveyor.setError(f'Failed to compute MDs')
if not self.conveyor.getError():
# run apply object, in charge of generate a prediction from idata
try:
sapply = sapply_child.SapplyChild(self.param, self.conveyor)
except:
LOG.warning(
'Sapply child architecture mismatch, defaulting to Sapply parent'
)
sapply = Sapply(self.param, self.conveyor)
sapply.run(cutoff, numsel, metric)
LOG.debug(
f'sapply child {type(sapply).__name__} completed `run()`')
# run odata object, in charge of formatting the prediction results
# note that if any of the above steps failed, an error has been inserted in the
# conveyor and odata will take case of showing an error message
try:
odata = odata_child.OdataChild(self.param, self.conveyor,
self.label)
except:
LOG.warning(
'Odata child architecture mismatch, defaulting to Odata parent'
)
odata = Odata(self.param, self.conveyor, self.label)
return odata.run()
class Build:
def __init__(self, model, param_file=None, param_string=None, output_format=None):
LOG.debug('Starting build...')
self.model = model
self.param = Parameters()
self.conveyor = Conveyor()
# identify the workflow type
self.conveyor.setOrigin('learn')
# generate a unique modelID
self.conveyor.addMeta('modelID',utils.id_generator())
LOG.debug(f'Generated new model with modelID: {self.conveyor.getMeta("modelID")}')
# load parameters
if param_file is not None:
# use the param_file to update existing parameters at the model
# directory and save changes to make them persistent
success, message = self.param.delta(model, 0, param_file, iformat='YAML')
elif param_string is not None:
success, message = self.param.delta(model, 0, param_string, iformat='JSONS')
else:
# load parameter file at the model directory
success, message = self.param.loadYaml(model, 0)
# being unable to load parameters is a critical error
if not success:
LOG.critical(f'Unable to load model parameters. {message}. Aborting...')
sys.exit(1)
# add additional output formats included in the constructor
# this is requiered to add JSON format as output when the object is
# instantiated from a web service call, requiring this output
if output_format is not None:
if output_format not in self.param.getVal('output_format'):
self.param.appVal('output_format',output_format)
if self.param.getVal('confidential'):
self.confidentialAuditParam()
def confidentialAuditParam (self):
import yaml
original_method = self.param.getVal('model')
if self.param.getVal ('quantitative'):
if original_method != 'PLSR':
self.param.setVal('model', 'PLSR')
LOG.info (f'CONFIDENTIALITY AUDIT: the model was set to PLSR, '
f'the original method {original_method} was not suitable to build confidential models')
else:
if original_method != 'PLSDA':
self.param.setVal('model', 'PLSDA')
LOG.info (f'CONFIDENTIALITY AUDIT: the model was set to PLSDA, '
f'the original method {original_method} was not suitable to build confidential models')
# TODO: conformal support
if self.param.getVal('conformal'):
self.param.setVal('conformal', False)
LOG.info ('CONFIDENTIALITY AUDIT: conformal was set to False. '
'Conformal models are not supported for now in confidential models')
parameters_file_path = utils.model_path(self.model, 0)
parameters_file_name = os.path.join (parameters_file_path,
'parameters.yaml')
with open(parameters_file_name, 'w') as pfile:
yaml.dump (self.param.p, pfile)
def get_ensemble(self):
''' Returns a Boolean indicating if the model uses external input
sources and a list with these sources '''
return self.param.getEnsemble()
def extend_modelID (self, ensembleID):
modelID = self.conveyor.getMeta('modelID')
modelID = f'{modelID}-{ensembleID}'
self.conveyor.addMeta('modelID', modelID)
LOG.debug (f'modelID re-defined as {self.conveyor.getVal("modelID")}')
def set_single_CPU(self) -> None:
''' Forces the use of a single CPU '''
LOG.debug('parameter "numCPUs" forced to be 1')
self.param.setVal('numCPUs',1)
def run(self, input_source):
''' Executes a default predicton workflow '''
# path to endpoint
epd = utils.model_path(self.model, 0)
# if not os.path.isdir(epd):
# self.conveyor.setError(f'Unable to find model {self.model}')
# #LOG.error(f'Unable to find model {self.model}')
# import ichild classes
# if not self.conveyor.getError():
# uses the child classes within the 'model' folder,
# to allow customization of the processing applied to each model
modpath = utils.module_path(self.model, 0)
idata_child = importlib.import_module(modpath+".idata_child")
learn_child = importlib.import_module(modpath+".learn_child")
odata_child = importlib.import_module(modpath+".odata_child")
# run idata object, in charge of generate model data from input
try:
idata = idata_child.IdataChild(self.param, self.conveyor, input_source)
except:
LOG.warning ('Idata child architecture mismatch, defaulting to Idata parent')
idata = Idata(self.param, self.conveyor, input_source)
idata.run()
LOG.debug(f'idata child {type(idata).__name__} completed `run()`')
if not self.conveyor.getError():
success, results = idata.preprocess_create()
if not success:
self.conveyor.setError(results)
if not self.conveyor.getError():
# check there is a suitable X and Y
if not self.conveyor.isKey ('xmatrix'):
self.conveyor.setError(f'Failed to compute MDs')
if not self.conveyor.isKey ('ymatrix'):
self.conveyor.setError(f'No activity data (Y) found in training series')
# run optional chemical space building for supporting "closest" training series object
# if self.param.getVal('buildSimilarity'):
if self.param.getVal('output_similar') is True:
from flame.slearn import Slearn
slearn_child = importlib.import_module(modpath+".slearn_child")
if not self.conveyor.getError():
# instantiate learn (build a space from idata) and run it
try:
slearn = slearn_child.SlearnChild(self.param, self.conveyor)
except:
LOG.warning ('Slearn child architecture mismatch, defaulting to Learn parent')
slearn = Slearn(self.param, self.conveyor)
slearn.run()
LOG.debug(f'slearn child {type(slearn).__name__} completed `run()`')
if not self.conveyor.getError():
# instantiate learn (build a model from idata) and run it
try:
learn = learn_child.LearnChild(self.param, self.conveyor)
except:
LOG.warning ('Learn child architecture mismatch, defaulting to Learn parent')
learn = Learn(self.param, self.conveyor)
learn.run()
LOG.debug(f'learn child {type(learn).__name__} completed `run()`')
# run odata object, in charge of formatting the prediction results
# note that if any of the above steps failed, an error has been inserted in the
# conveyor and odata will take case of showing an error message
try:
odata = odata_child.OdataChild(self.param, self.conveyor)
except:
LOG.warning ('Odata child architecture mismatch, defaulting to Odata parent')
odata = Odata(self.param, self.conveyor)
return odata.run()
def action_parameters(model, version=None, oformat='text'):
''' Returns an object with whole results info for a given model and version '''
if model is None:
return False, 'Empty model label'
from flame.parameters import Parameters
param = Parameters()
success, results = param.loadYaml(model, version)
if not success:
print(f'error obtaining parametes for model {model} : {results}')
return False, results
if oformat != 'text':
return True, param
else:
yaml = param.dumpYAML()
for line in yaml:
print(line)
# order = ['input_type', 'quantitative', 'SDFile_activity', 'SDFile_name', 'SDFile_id',
# 'SDFile_experimental', 'SDFile_complementary', 'normalize_method', 'ionize_method', 'convert3D_method',
# 'computeMD_method', 'model', 'modelAutoscaling', 'tune', 'conformal',
# 'conformalConfidence', 'ModelValidationCV', 'ModelValidationLC',
# 'ModelValidationN', 'ModelValidationP', 'output_format', 'output_md', 'output_similar',
# 'TSV_activity', 'TSV_objnames', 'TSV_varnames', 'imbalance',
# 'feature_selection', 'feature_number', 'mol_batch',
# 'ensemble_names','ensemble_versions',
# 'similarity_metric', 'similarity_cutoff_num', 'similarity_cutoff_distance',
# 'numCPUs', 'verbose_error', 'modelingToolkit',
# 'endpoint', 'model_path',
# #'md5',
# 'version']
# order += ['MD_settings', 'RF_parameters','RF_optimize',
# 'SVM_parameters','SVM_optimize',
# 'PLSDA_parameters','PLSDA_optimize',
# 'PLSR_parameters','PLSR_optimize',
# 'GNB_parameters']
# for ik in order:
# if ik in param.p:
# k = ik
# v = param.p[k]
# ivalue = ''
# idescr = ''
# ioptio = ''
# ## newest parameter formats are extended and contain
# ## rich metainformation for each entry
# if param.extended:
# if 'value' in v:
# if not isinstance(v['value'] ,dict):
# ivalue = v['value']
# else:
# # print header of dictionaty
# print (f'{k} :')
# # iterate keys assuming existence of value and description
# for intk in v['value']:
# intv = v['value'][intk]
# iivalue = ''
# if "value" in intv:
# iivalue = intv["value"]
# iidescr = ''
# if "description" in intv and intv["description"] is not None:
# iidescr = intv["description"]
# iioptio = ''
# if 'options' in intv:
# toptio = intv['options']
# if isinstance(toptio, list):
# if toptio != [None]:
# iioptio = f' {toptio}'
# if isinstance (iivalue, float):
# iivalue = f'{iivalue:f}'
# elif iivalue is None:
# iivalue = ''
# print (f' {intk:27} : {str(iivalue):30} #{iioptio} {iidescr}')
# continue
# if 'description' in v:
# idescr = v['description']
# if 'options' in v:
# toptio = v['options']
# if isinstance(toptio, list):
# ioptio = f' {toptio}'
# ### compatibility: old stile parameters
# else:
# if not isinstance(v ,dict):
# ivalue = v
# else:
# ivalue = '*dictionary*'
# ### end compatibility
# if isinstance (ivalue, float):
# ivalue = f'{ivalue:f}'
# elif ivalue is None:
# ivalue = ''
# print (f'{k:30} : {str(ivalue):30} #{ioptio} {idescr}')
return True, 'parameters listed'