From 2e6d1c0420cf7ce3a452025c613e629409d8931d Mon Sep 17 00:00:00 2001
From: Matt Aton <matthewaton2@gmail.com>
Date: Thu, 5 Dec 2024 09:50:30 -0700
Subject: [PATCH] Speedup import time with lazy imports (#987)

* Initial commit

* Implement simple solution

* Remove comments and spaces

* Fix linting

* Update Changelog
---
 ChangeLog.md  | 4 ++++
 biom/table.py | 8 +++++---
 2 files changed, 9 insertions(+), 3 deletions(-)

diff --git a/ChangeLog.md b/ChangeLog.md
index 3e28019d..b2f048e6 100644
--- a/ChangeLog.md
+++ b/ChangeLog.md
@@ -4,6 +4,10 @@ BIOM-Format ChangeLog
 biom-2.1.16-dev
 ---------------
 
+Performance improvements:
+
+* Decreased execution time of `import biom` by half with lazy imports. See PR[#987](https://github.com/biocore/biom-format/pull/987)
+
 Maintenance:
 
 * Python 3.7 and 3.8 removed from CI as they are [end-of-life](https://devguide.python.org/versions/). Python 3.13 added to CI. See PR[#986](https://github.com/biocore/biom-format/pull/986).
diff --git a/biom/table.py b/biom/table.py
index 4b56ab56..2cd40967 100644
--- a/biom/table.py
+++ b/biom/table.py
@@ -172,8 +172,6 @@
 # -----------------------------------------------------------------------------
 
 import numpy as np
-import scipy.stats
-import h5py
 from copy import deepcopy
 from datetime import datetime
 from json import dumps as _json_dumps, JSONEncoder
@@ -184,7 +182,6 @@
 from numpy import ndarray, asarray, zeros, newaxis
 from scipy.sparse import (coo_matrix, csc_matrix, csr_matrix, isspmatrix,
                           vstack, hstack, dok_matrix)
-import pandas as pd
 import re
 from biom.exception import (TableException, UnknownAxisError, UnknownIDError,
                             DisjointIDError)
@@ -3234,6 +3231,8 @@ def rankdata(self, axis='sample', inplace=True, method='average'):
         o4	1.0	4.0	1.0
 
         """
+        import scipy.stats
+
         def f(val, id_, _):
             return scipy.stats.rankdata(val, method=method)
         return self.transform(f, axis=axis, inplace=inplace)
@@ -4108,6 +4107,7 @@ def from_hdf5(cls, h5grp, ids=None, axis='sample', parse_fs=None,
         >>>     t = Table.from_hdf5(f, ids=["GG_OTU_1"],
         ...                         axis='observation') # doctest: +SKIP
         """
+        import h5py
         if not isinstance(h5grp, (h5py.Group, h5py.File)):
             raise ValueError("h5grp does not appear to be an HDF5 file or "
                              "group")
@@ -4341,6 +4341,7 @@ def to_dataframe(self, dense=False):
         index = self.ids(axis='observation')
         columns = self.ids()
 
+        import pandas as pd
         if dense:
             mat = self.matrix_data.toarray()
             constructor = pd.DataFrame
@@ -4442,6 +4443,7 @@ def metadata_to_dataframe(self, axis):
         O1   Bacteria     Firmicutes
         O2   Bacteria  Bacteroidetes
         """
+        import pandas as pd
         md = self.metadata(axis=axis)
         if md is None:
             raise KeyError("%s does not have metadata" % axis)