pertpy 0.6.0__py3-none-any.whl → 0.8.0__py3-none-any.whl

pertpy/__init__.py

@@ -2,7 +2,7 @@
 
 __author__ = "Lukas Heumos"
 __email__ = "lukas.heumos@posteo.net"
-__version__ = "0.6.0"
+__version__ = "0.8.0"
 
 import warnings
 
@@ -11,9 +11,11 @@ from numba import NumbaDeprecationWarning
 
 warnings.filterwarnings("ignore", category=NumbaDeprecationWarning)
 warnings.filterwarnings("ignore", category=MatplotlibDeprecationWarning)
-warnings.filterwarnings("ignore", category=UserWarning)
+warnings.filterwarnings("ignore", category=SyntaxWarning)
+warnings.filterwarnings("ignore", category=UserWarning, module="scvi._settings")
 
 from . import data as dt
+from . import metadata as md
 from . import plot as pl
 from . import preprocessing as pp
 from . import tools as tl

pertpy/data/__init__.py

@@ -7,12 +7,14 @@ from pertpy.data._datasets import (
     burczynski_crohn,
     chang_2021,
     cinemaot_example,
+    combosciplex,
     datlinger_2017,
     datlinger_2021,
     dialogue_example,
     distance_example,
     dixit_2016,
     dixit_2016_raw,
+    dong_2023,
     frangieh_2021,
     frangieh_2021_protein,
     frangieh_2021_raw,
@@ -22,6 +24,7 @@ from pertpy.data._datasets import (
     gasperini_2019_lowmoi,
     gehring_2019,
     haber_2017_regions,
+    hagai_2018,
     kang_2018,
     mcfarland_2020,
     norman_2019,
@@ -36,17 +39,79 @@ from pertpy.data._datasets import (
     schraivogel_2020_tap_screen_chr8,
     schraivogel_2020_tap_screen_chr11,
     sciplex3_raw,
+    sciplex_gxe1,
     shifrut_2018,
-    smillie,
+    smillie_2019,
     srivatsan_2020_sciplex2,
     srivatsan_2020_sciplex3,
     srivatsan_2020_sciplex4,
     stephenson_2021_subsampled,
+    tasccoda_example,
     tian_2019_day7neuron,
     tian_2019_ipsc,
     tian_2021_crispra,
     tian_2021_crispri,
     weinreb_2020,
     xie_2017,
+    zhang_2021,
     zhao_2021,
 )
+
+__all__ = [
+    "adamson_2016_pilot",
+    "adamson_2016_upr_epistasis",
+    "adamson_2016_upr_perturb_seq",
+    "aissa_2021",
+    "bhattacherjee",
+    "burczynski_crohn",
+    "chang_2021",
+    "cinemaot_example",
+    "combosciplex",
+    "datlinger_2017",
+    "datlinger_2021",
+    "dialogue_example",
+    "distance_example",
+    "dixit_2016",
+    "dixit_2016_raw",
+    "dong_2023",
+    "frangieh_2021",
+    "frangieh_2021_protein",
+    "frangieh_2021_raw",
+    "frangieh_2021_rna",
+    "gasperini_2019_atscale",
+    "gasperini_2019_highmoi",
+    "gasperini_2019_lowmoi",
+    "gehring_2019",
+    "haber_2017_regions",
+    "hagai_2018",
+    "kang_2018",
+    "mcfarland_2020",
+    "norman_2019",
+    "norman_2019_raw",
+    "papalexi_2021",
+    "replogle_2022_k562_essential",
+    "replogle_2022_k562_gwps",
+    "replogle_2022_rpe1",
+    "sc_sim_augur",
+    "schiebinger_2019_16day",
+    "schiebinger_2019_18day",
+    "schraivogel_2020_tap_screen_chr8",
+    "schraivogel_2020_tap_screen_chr11",
+    "sciplex3_raw",
+    "sciplex_gxe1",
+    "shifrut_2018",
+    "smillie_2019",
+    "srivatsan_2020_sciplex2",
+    "srivatsan_2020_sciplex3",
+    "srivatsan_2020_sciplex4",
+    "stephenson_2021_subsampled",
+    "tasccoda_example",
+    "tian_2019_day7neuron",
+    "tian_2019_ipsc",
+    "tian_2021_crispra",
+    "tian_2021_crispri",
+    "weinreb_2020",
+    "xie_2017",
+    "zhao_2021",
+    "zhang_2021",
+]

pertpy/data/_dataloader.py

@@ -2,18 +2,18 @@ import tempfile
 from pathlib import Path
 from random import choice
 from string import ascii_lowercase
-from typing import Union
 from zipfile import ZipFile
 
 import requests
-from rich import print
+from filelock import FileLock
+from lamin_utils import logger
 from rich.progress import Progress
 
 
 def _download(  # pragma: no cover
     url: str,
     output_file_name: str = None,
-    output_path: Union[str, Path] = None,
+    output_path: str | Path = None,
     block_size: int = 1024,
     overwrite: bool = False,
     is_zip: bool = False,
@@ -38,31 +38,33 @@ def _download(  # pragma: no cover
     download_to_path = (
         f"{output_path}{output_file_name}" if str(output_path).endswith("/") else f"{output_path}/{output_file_name}"
     )
-    if Path(download_to_path).exists():
-        warning = f"[bold red]File {download_to_path} already exists!"
-        if not overwrite:
-            print(warning)
+
+    Path(output_path).mkdir(parents=True, exist_ok=True)
+    lock_path = f"{output_path}/{output_file_name}.lock"
+    with FileLock(lock_path):
+        if Path(download_to_path).exists() and not overwrite:
+            logger.warning(f"File {download_to_path} already exists!")
             return
-        else:
-            print(f"{warning} Overwriting...")
 
-    response = requests.get(url, stream=True)
-    total = int(response.headers.get("content-length", 0))
+        temp_file_name = f"{download_to_path}.part"
+
+        response = requests.get(url, stream=True)
+        total = int(response.headers.get("content-length", 0))
+
+        with Progress(refresh_per_second=100) as progress:
+            task = progress.add_task("[red]Downloading...", total=total)
+            with Path(temp_file_name).open("wb") as file:
+                for data in response.iter_content(block_size):
+                    file.write(data)
+                    progress.update(task, advance=block_size)
+            progress.update(task, completed=total, refresh=True)
 
-    with Progress(refresh_per_second=100) as progress:
-        task = progress.add_task("[red]Downloading...", total=total)
-        Path(output_path).mkdir(parents=True, exist_ok=True)
-        with Path(download_to_path).open("wb") as file:
-            for data in response.iter_content(block_size):
-                file.write(data)
-                progress.update(task, advance=block_size)
+        Path(temp_file_name).replace(download_to_path)
 
-        # force the progress bar to 100% at the end
-        progress.update(task, completed=total, refresh=True)
+        if is_zip:
+            output_path = output_path or tempfile.gettempdir()
+            with ZipFile(download_to_path, "r") as zip_obj:
+                zip_obj.extractall(path=output_path)
+                zip_obj.namelist()
 
-    if is_zip:
-        output_path = output_path or tempfile.gettempdir()
-        with ZipFile(download_to_path, "r") as zip_obj:
-            zip_obj.extractall(path=output_path)
-            extracted = zip_obj.namelist()
-            print(extracted)
+    Path(lock_path).unlink()