def load_and_deserialize_plots(path):
"""Returns deserialized plots loaded from given path."""
result = []
for record in tf.python_io.tf_record_iterator(path):
plots_for_slice = metrics_for_slice_pb2.PlotsForSlice.FromString(
record)
if plots_for_slice.HasField('plot_data'):
result.append((
slicer.deserialize_slice_key(plots_for_slice.slice_key), # pytype: disable=wrong-arg-types
plots_for_slice.plot_data))
if plots_for_slice.plots:
result.append((
slicer.deserialize_slice_key(plots_for_slice.slice_key), # pytype: disable=wrong-arg-types
plots_for_slice.plots))
return result
def load_and_deserialize_plots(
path: Text) -> List[Tuple[slicer.SliceKeyType, Any]]:
"""Returns deserialized plots loaded from given path."""
result = []
for record in tf.compat.v1.python_io.tf_record_iterator(path):
plots_for_slice = metrics_for_slice_pb2.PlotsForSlice.FromString(
record)
plots_map = {}
if plots_for_slice.plots:
plot_dict = _convert_proto_map_to_dict(plots_for_slice.plots)
keys = list(plot_dict.keys())
# If there is only one label, choose it automatically.
plot_data = plot_dict[keys[0]] if len(keys) == 1 else plot_dict
plots_map[''] = {'': plot_data}
elif plots_for_slice.HasField('plot_data'):
plots_map[''] = {
'': json_format.MessageToDict(plots_for_slice.plot_data)
}
if plots_for_slice.plot_keys_and_values:
for kv in plots_for_slice.plot_keys_and_values:
output_name = kv.key.output_name
if output_name not in plots_map:
plots_map[output_name] = {}
sub_key_id = _get_sub_key_id(
kv.key.sub_key) if kv.key.HasField('sub_key') else ''
plots_map[output_name][sub_key_id] = json_format.MessageToDict(
kv.value)
result.append((
slicer.deserialize_slice_key(plots_for_slice.slice_key), # pytype: disable=wrong-arg-types
plots_map))
return result
def load_and_deserialize_metrics(
path: Text,
model_name: Optional[Text] = None
) -> List[Tuple[slicer.SliceKeyType, Any]]:
"""Loads metrics from the given location and builds a metric map for it."""
result = []
for record in tf.compat.v1.python_io.tf_record_iterator(path):
metrics_for_slice = metrics_for_slice_pb2.MetricsForSlice.FromString(
record)
model_metrics_map = {}
if metrics_for_slice.metrics:
model_metrics_map[''] = {
'': {
'': _convert_proto_map_to_dict(metrics_for_slice.metrics)
}
}
if metrics_for_slice.metric_keys:
for key, value in zip(metrics_for_slice.metric_keys,
metrics_for_slice.metric_values):
current_model_name = key.model_name
if current_model_name not in model_metrics_map:
model_metrics_map[current_model_name] = {}
output_name = key.output_name
if output_name not in model_metrics_map[current_model_name]:
model_metrics_map[current_model_name][output_name] = {}
multi_class_metrics_map = model_metrics_map[
current_model_name][output_name]
multi_class_key_id = _get_multi_class_key_id(
key.multi_class_key) if key.HasField(
'multi_class_key') else ''
if multi_class_key_id not in multi_class_metrics_map:
multi_class_metrics_map[multi_class_key_id] = {}
metric_name = key.name
multi_class_metrics_map[multi_class_key_id][
metric_name] = json_format.MessageToDict(value)
metrics_map = None
keys = list(model_metrics_map.keys())
if model_name in model_metrics_map:
# Use the provided model name if there is a match.
metrics_map = model_metrics_map[model_name]
elif model_name is None and len(keys) == 1:
# Show result of the only model if no model name is specified.
metrics_map = model_metrics_map[keys[0]]
else:
# No match found.
raise ValueError('Fail to find metrics for model name: %s . '
'Available model names are [%s]' %
(model_name, ', '.join(keys)))
result.append((
slicer.deserialize_slice_key(metrics_for_slice.slice_key), # pytype: disable=wrong-arg-types
metrics_map))
return result
def load_and_deserialize_metrics(
path):
result = []
for record in tf.python_io.tf_record_iterator(path):
metrics_for_slice = metrics_for_slice_pb2.MetricsForSlice.FromString(record)
result.append((
slicer.deserialize_slice_key(metrics_for_slice.slice_key), # pytype: disable=wrong-arg-types
metrics_for_slice.metrics))
return result
def load_and_deserialize_metrics(path: Text
) -> List[Tuple[slicer.SliceKeyType, Any]]:
result = []
for record in tf.compat.v1.python_io.tf_record_iterator(path):
metrics_for_slice = metrics_for_slice_pb2.MetricsForSlice.FromString(record)
result.append((
slicer.deserialize_slice_key(metrics_for_slice.slice_key), # pytype: disable=wrong-arg-types
metrics_for_slice.metrics))
return result
def load_and_deserialize_plots(
path: Text) -> List[Tuple[slicer.SliceKeyType, Any]]:
"""Returns deserialized plots loaded from given path."""
result = []
for record in tf.python_io.tf_record_iterator(path):
plots_for_slice = metrics_for_slice_pb2.PlotsForSlice.FromString(record)
if plots_for_slice.HasField('plot_data'):
if plots_for_slice.plots:
raise RuntimeError('Both plots and plot_data are set.')
# For backward compatibility, plots data geneated with old code are added
# to the plots map with default key empty string.
plots_for_slice.plots[''].CopyFrom(plots_for_slice.plot_data)
result.append((
slicer.deserialize_slice_key(plots_for_slice.slice_key), # pytype: disable=wrong-arg-types
plots_for_slice.plots))
return result
def testDeserializeSliceKey(self):
slice_metrics = text_format.Parse(
"""
single_slice_keys {
column: 'age'
int64_value: 5
}
single_slice_keys {
column: 'language'
bytes_value: 'english'
}
single_slice_keys {
column: 'price'
float_value: 1.0
}
""", metrics_for_slice_pb2.SliceKey())
got_slice_key = slicer.deserialize_slice_key(slice_metrics)
self.assertItemsEqual([('age', 5), ('language', b'english'),
('price', 1.0)], got_slice_key)
