ClearML/clearml
1
0
Fork 0
mirror of https://github.com/clearml/clearml synced 2025-03-03 18:52:12 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Fix media preview is created instead of a table preview

Browse Source
This commit is contained in:
allegroai 2022-11-11 15:34:43 +02:00
parent 251cfa30a9
commit b58c70f65e
1 changed files with 16 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
clearml/datasets/dataset.py
View File

@ -2751,6 +2751,13 @@ class Dataset(object):
nrows=self.__preview_tabular_row_count, nrows=self.__preview_tabular_row_count,
compression=compression_.lstrip(".") if compression_ else None, compression=compression_.lstrip(".") if compression_ else None,
) )
elif file_extension_ == ".tsv" and pd:
return pd.read_csv(
file_path_,
sep='\t',
nrows=self.__preview_tabular_row_count,
compression=compression_.lstrip(".") if compression_ else None,
)
elif file_extension_ == ".parquet" or file_extension_ == ".parq": elif file_extension_ == ".parquet" or file_extension_ == ".parq":
if pyarrow: if pyarrow:
pf = pyarrow.parquet.ParquetFile(file_path_) pf = pyarrow.parquet.ParquetFile(file_path_)
@ -2765,7 +2772,7 @@ class Dataset(object):
return None return None
compression_extensions = {".gz", ".bz2", ".zip", ".xz", ".zst"} compression_extensions = {".gz", ".bz2", ".zip", ".xz", ".zst"}
tabular_extensions = {".csv", ".parquet", ".parq", ".npz", ".npy"} tabular_extensions = {".csv", ".tsv", ".parquet", ".parq", ".npz", ".npy"}
for file in self._dataset_file_entries.values(): for file in self._dataset_file_entries.values():
if file.local_path: if file.local_path:
file_path = file.local_path file_path = file.local_path
@ -2786,9 +2793,14 @@ class Dataset(object):
if artifact is not None: if artifact is not None:
# noinspection PyBroadException # noinspection PyBroadException
try: try:
self._task.get_logger().report_media( if isinstance(artifact, pd.DataFrame):
"Tables", file_name, stream=artifact.to_csv(index=False), file_extension=".txt" self._task.get_logger().report_table(
) "Tables", "summary", table_plot=artifact
)
else:
self._task.get_logger().report_media(
"Tables", file_name, stream=artifact.to_csv(index=False), file_extension=".txt"
)
self.__preview_tables_count += 1 self.__preview_tables_count += 1
except Exception: except Exception:
pass pass