From 4536d6ef8a35e49f224ec5e2540abd9e1ab0fe6d Mon Sep 17 00:00:00 2001
From: Nathan Thorpe <nathan@techstormpc.com>
Date: Thu, 24 Jul 2025 14:28:39 -0700
Subject: [PATCH 1/2] lazy load pandas

---
 cirro/cli/controller.py             | 3 ++-
 cirro/helpers/preprocess_dataset.py | 9 ++++++---
 cirro/sdk/file.py                   | 8 ++++----
 3 files changed, 12 insertions(+), 8 deletions(-)

diff --git a/cirro/cli/controller.py b/cirro/cli/controller.py
index e5ad6ed3..d586e9a4 100644
--- a/cirro/cli/controller.py
+++ b/cirro/cli/controller.py
@@ -5,7 +5,6 @@
 import sys
 from pathlib import Path
 
-import pandas as pd
 import requests
 from cirro_api_client.v1.models import UploadDatasetRequest, Status, Executor
 
@@ -58,6 +57,8 @@ def run_list_datasets(input_params: ListArguments, interactive=False):
     datasets = cirro.datasets.list(input_params['project'])
 
     sorted_datasets = sorted(datasets, key=lambda d: d.created_at, reverse=True)
+
+    import pandas as pd
     df = pd.DataFrame.from_records([d.to_dict() for d in sorted_datasets])
     df = df[['id', 'name', 'description', 'processId', 'status', 'createdBy', 'createdAt']]
     print(df.to_string())
diff --git a/cirro/helpers/preprocess_dataset.py b/cirro/helpers/preprocess_dataset.py
index 7a80e6e6..09d2238c 100644
--- a/cirro/helpers/preprocess_dataset.py
+++ b/cirro/helpers/preprocess_dataset.py
@@ -2,12 +2,14 @@
 import logging
 import os
 from pathlib import Path
+from typing import TYPE_CHECKING
 
 import boto3
-import pandas as pd
 
-from cirro.models.s3_path import S3Path
+if TYPE_CHECKING:
+    from pandas import DataFrame
 
+from cirro.models.s3_path import S3Path
 
 def _write_json(dat, local_path: str, indent=4):
     with Path(local_path).open(mode="wt") as handle:
@@ -66,11 +68,12 @@ def log(self):
         self.logger.info(f"Number of files in dataset: {self.files.shape[0]:,}")
         self.logger.info(f"Number of samples in dataset: {self.samplesheet.shape[0]:,}")
 
-    def _read_csv(self, suffix: str, required_columns=None) -> pd.DataFrame:
+    def _read_csv(self, suffix: str, required_columns=None) -> 'DataFrame':
         """Read a CSV from the dataset and check for any required columns."""
         if required_columns is None:
             required_columns = []
 
+        import pandas as pd
         df = pd.read_csv(f"{self.s3_dataset}/{suffix}")
         for col in required_columns:
             assert col in df.columns.values, f"Did not find expected columns {col} in {self.s3_dataset}/{suffix}"
diff --git a/cirro/sdk/file.py b/cirro/sdk/file.py
index 91b9b27e..35b70e42 100644
--- a/cirro/sdk/file.py
+++ b/cirro/sdk/file.py
@@ -2,11 +2,10 @@
 from io import BytesIO, StringIO
 from typing import List
 
-import pandas as pd
-
 from typing import TYPE_CHECKING
 if TYPE_CHECKING:
     import anndata
+    from pandas import DataFrame
 
 from cirro.cirro_client import CirroApi
 from cirro.models.file import File
@@ -88,7 +87,7 @@ def _get(self) -> bytes:
 
         return self._client.file.get_file(self._file)
 
-    def read_csv(self, compression='infer', encoding='utf-8', **kwargs) -> pd.DataFrame:
+    def read_csv(self, compression='infer', encoding='utf-8', **kwargs) -> 'DataFrame':
         """
         Parse the file as a Pandas DataFrame.
 
@@ -100,6 +99,7 @@ def read_csv(self, compression='infer', encoding='utf-8', **kwargs) -> pd.DataFr
         All other keyword arguments are passed to pandas.read_csv
         https://pandas.pydata.org/docs/reference/api/pandas.read_csv.html
         """
+        import pandas
 
         if compression == 'infer':
             # If the file appears to be compressed
@@ -119,7 +119,7 @@ def read_csv(self, compression='infer', encoding='utf-8', **kwargs) -> pd.DataFr
         else:
             handle = StringIO(self._get().decode(encoding))
 
-        df = pd.read_csv(
+        df = pandas.read_csv(
             handle,
             compression=compression,
             encoding=encoding,

From cc7b82027a5619cc6fad2cb8db1d1f237bc48f27 Mon Sep 17 00:00:00 2001
From: Nathan Thorpe <nathan@techstormpc.com>
Date: Thu, 24 Jul 2025 14:30:43 -0700
Subject: [PATCH 2/2] lint

---
 cirro/helpers/preprocess_dataset.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/cirro/helpers/preprocess_dataset.py b/cirro/helpers/preprocess_dataset.py
index 09d2238c..6436eb1c 100644
--- a/cirro/helpers/preprocess_dataset.py
+++ b/cirro/helpers/preprocess_dataset.py
@@ -11,6 +11,7 @@
 
 from cirro.models.s3_path import S3Path
 
+
 def _write_json(dat, local_path: str, indent=4):
     with Path(local_path).open(mode="wt") as handle:
         return json.dump(dat, handle, indent=indent)