dclab 0.63.0__tar.gz → 0.63.1__tar.gz

{dclab-0.63.0 → dclab-0.63.1}/CHANGELOG

@@ -1,3 +1,7 @@
+0.63.1
+ - docs: update documentation and code references for kde submodule
+ - enh: cache S3 sessions and clients (test suite duration reduced by 30%)
+ - enh: cache "basins" response for DCOR format
 0.63.0
  - feat: add `progress_callback` argument to all data export methods (#271)
 0.62.17

{dclab-0.63.0 → dclab-0.63.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dclab
-Version: 0.63.0
+Version: 0.63.1
 Summary: Library for real-time deformability cytometry (RT-DC)
 Author: Benedikt Hartmann, Eoghan O'Connell, Maik Herbig, Maximilian Schlögel, Nadia Sbaa, Paul Müller, Philipp Rosendahl, Raghava Alajangi
 Maintainer-email: Paul Müller <dev@craban.de>

{dclab-0.63.0 → dclab-0.63.1}/dclab/_version.py

@@ -17,5 +17,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.63.0'
-__version_tuple__ = version_tuple = (0, 63, 0)
+__version__ = version = '0.63.1'
+__version_tuple__ = version_tuple = (0, 63, 1)

{dclab-0.63.0 → dclab-0.63.1}/dclab/rtdc_dataset/fmt_dcor/api.py

@@ -12,7 +12,7 @@ class DCORAccessError(BaseException):
 class APIHandler:
     """Handles the DCOR api with caching for simple queries"""
     #: these are cached to minimize network usage
-    cache_queries = ["metadata", "size", "feature_list", "valid"]
+    cache_queries = ["basins", "metadata", "size", "feature_list", "valid"]
     #: DCOR API Keys/Tokens in the current session
     api_keys = []
 

{dclab-0.63.0 → dclab-0.63.1}/dclab/rtdc_dataset/fmt_s3.py

@@ -42,6 +42,27 @@ S3_ACCESS_KEY_ID = os.environ.get("DCLAB_S3_ACCESS_KEY_ID")
 S3_SECRET_ACCESS_KEY = os.environ.get("DCLAB_S3_SECRET_ACCESS_KEY")
 
 
+@functools.lru_cache(maxsize=1000)
+def get_s3_session_client(access_key_id: str,
+                          secret_access_key: str,
+                          use_ssl: bool,
+                          verify_ssl: bool,
+                          endpoint_url: str
+                          ):
+    botocore_session = botocore.session.get_session()
+    s3_session = boto3.Session(
+        aws_access_key_id=access_key_id,
+        aws_secret_access_key=secret_access_key,
+        botocore_session=botocore_session)
+    s3_client = s3_session.client(
+        service_name='s3',
+        use_ssl=use_ssl,
+        verify=verify_ssl,
+        endpoint_url=endpoint_url,
+    )
+    return botocore_session, s3_session, s3_client
+
+
 class S3File(HTTPFile):
     """Monkeypatched `HTTPFile` to support authenticated access to S3"""
     def __init__(self,
@@ -74,17 +95,15 @@ class S3File(HTTPFile):
                 "not specify the full S3 URL or that you forgot to set "
                 "the `S3_ENDPOINT_URL` environment variable.")
         endpoint_url = endpoint_url.strip().rstrip("/")
-        self.botocore_session = botocore.session.get_session()
-        self.s3_session = boto3.Session(
-            aws_access_key_id=access_key_id,
-            aws_secret_access_key=secret_access_key,
-            botocore_session=self.botocore_session)
-        self.s3_client = self.s3_session.client(
-            service_name='s3',
-            use_ssl=use_ssl,
-            verify=verify_ssl,
-            endpoint_url=endpoint_url,
+        self.botocore_session, self.s3_session, self.s3_client = \
+            get_s3_session_client(
+                access_key_id=access_key_id,
+                secret_access_key=secret_access_key,
+                use_ssl=use_ssl,
+                verify_ssl=verify_ssl,
+                endpoint_url=endpoint_url,
             )
+
         # Use a configuration that allows anonymous access
         # https://stackoverflow.com/a/34866092
         if not secret_access_key:

{dclab-0.63.0 → dclab-0.63.1}/dclab.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dclab
-Version: 0.63.0
+Version: 0.63.1
 Summary: Library for real-time deformability cytometry (RT-DC)
 Author: Benedikt Hartmann, Eoghan O'Connell, Maik Herbig, Maximilian Schlögel, Nadia Sbaa, Paul Müller, Philipp Rosendahl, Raghava Alajangi
 Maintainer-email: Paul Müller <dev@craban.de>

{dclab-0.63.0 → dclab-0.63.1}/docs/sec_av_scatter.rst

@@ -5,24 +5,28 @@ Scatter plots
 =============
 
 For data visualization, dclab comes with predefined 
-:ref:`kernel density estimators (KDEs) <sec_ref_kde_methods>` and
+:ref:`kernel density estimators (KDEs) <sec_ref_kde>` and
 an :ref:`event downsampling <sec_ref_downsampling>` module.
 The functionalities of both modules are made available directly via the
-:class:`RTDCBase <dclab.rtdc_dataset.RTDCBase>` class.
+:py:class:`dclab.kde.KernelDensityEstimator` class. 
 
 KDE scatter plot
 ----------------
 The KDE of the events in a 2D scatter plot can be used to
 colorize events according to event density using the
-:func:`RTDCBase.get_kde_scatter <dclab.rtdc_dataset.RTDCBase.get_kde_scatter>`
+:py:meth:`~dclab.kde.KernelDensityEstimator.get_scatter`
 function.
 
 .. plot::
 
     import matplotlib.pylab as plt
     import dclab
+    from dclab.kde import KernelDensityEstimator
+    # load the example dataset
     ds = dclab.new_dataset("data/example.rtdc")
-    kde = ds.get_kde_scatter(xax="area_um", yax="deform")
+    # create a kernel density estimator
+    kde_instance = KernelDensityEstimator(ds)
+    kde = kde_instance.get_scatter(xax="area_um", yax="deform")
     
     ax = plt.subplot(111, title="scatter plot with {} events".format(len(kde)))
     sc = ax.scatter(ds["area_um"], ds["deform"], c=kde, marker=".")
@@ -65,8 +69,8 @@ Frequently, data is visualized on logarithmic scales. If the KDE
 is computed on a linear scale, then the result will look unaesthetic
 when plotted on a logarithmic scale. Therefore, the methods
 :func:`get_downsampled_scatter <dclab.rtdc_dataset.RTDCBase.get_downsampled_scatter>`,
-:func:`get_kde_contour <dclab.rtdc_dataset.RTDCBase.get_kde_contour>`, and
-:func:`get_kde_scatter <dclab.rtdc_dataset.RTDCBase.get_kde_scatter>`
+:py:meth:`~dclab.kde.KernelDensityEstimator.get_contour`, and
+:py:meth:`~dclab.kde.KernelDensityEstimator.get_scatter`
 offer the keyword arguments ``xscale`` and ``yscale`` which can be set to
 "log" for prettier plots.
 
@@ -74,9 +78,13 @@ offer the keyword arguments ``xscale`` and ``yscale`` which can be set to
 
     import matplotlib.pylab as plt
     import dclab
+    from dclab.kde import KernelDensityEstimator
+    # load the example dataset
     ds = dclab.new_dataset("data/example.rtdc")
-    kde_lin = ds.get_kde_scatter(xax="area_um", yax="deform", yscale="linear")
-    kde_log = ds.get_kde_scatter(xax="area_um", yax="deform", yscale="log")
+    # create a kernel density estimator
+    kde_instance = KernelDensityEstimator(ds)
+    kde_lin = kde_instance.get_scatter(xax="area_um", yax="deform", yscale="linear")
+    kde_log = kde_instance.get_scatter(xax="area_um", yax="deform", yscale="log")
 
     ax1 = plt.subplot(121, title="KDE with linear y-scale")
     sc1 = ax1.scatter(ds["area_um"], ds["deform"], c=kde_lin, marker=".")
@@ -105,8 +113,11 @@ Isoelasticity lines are available via the
 
     import matplotlib.pylab as plt
     import dclab
+    from dclab.kde import KernelDensityEstimator
+    # load the example dataset
     ds = dclab.new_dataset("data/example.rtdc")
-    kde = ds.get_kde_scatter(xax="area_um", yax="deform")
+    kde_instance = KernelDensityEstimator(ds)
+    kde = kde_instance.get_scatter(xax="area_um", yax="deform")
 
     isodef = dclab.isoelastics.get_default()
     iso = isodef.get_with_rtdcbase(method="numerical",
@@ -131,20 +142,23 @@ Contour plot with percentiles
 Contour plots are commonly used to compare the kernel density
 between measurements. Kernel density estimates (on a grid) for contour
 plots can be computed with the function
-:func:`RTDCBase.get_kde_contour <dclab.rtdc_dataset.RTDCBase.get_kde_contour>`.
+:py:meth:`~dclab.kde.KernelDensityEstimator.get_contour`.
 In addition, it is possible to compute contours at data
 `percentiles <https://en.wikipedia.org/wiki/Percentile>`_
-using :func:`dclab.kde_contours.get_quantile_levels`.
+using :func:`~dclab.kde.contours.get_quantile_levels`.
 
 .. plot::
 
     import matplotlib.pylab as plt
     import dclab
+    from dclab.kde import KernelDensityEstimator
+    # load the example dataset
     ds = dclab.new_dataset("data/example.rtdc")
-    X, Y, Z = ds.get_kde_contour(xax="area_um", yax="deform")
+    kde_instance = KernelDensityEstimator(ds)
+    X, Y, Z = kde_instance.get_contour(xax="area_um", yax="deform")
     Z /= Z.max()
     quantiles = [.1, .5, .75]
-    levels = dclab.kde_contours.get_quantile_levels(density=Z,
+    levels = dclab.kde.contours.get_quantile_levels(density=Z,
                                                     x=X,
                                                     y=Y,
                                                     xp=ds["area_um"],
@@ -173,7 +187,7 @@ using :func:`dclab.kde_contours.get_quantile_levels`.
     plt.show()
 
 Note that you may compute (and plot) the contour lines directly
-yourself using the function :func:`dclab.kde_contours.find_contours_level`.
+yourself using the function :func:`~dclab.kde.contours.find_contours_level`.
 
 
 
@@ -189,8 +203,12 @@ in Shape-Out and then import them in dclab.
 
     import matplotlib.pylab as plt
     import dclab
+    from dclab.kde import KernelDensityEstimator
+    # load the example dataset
     ds = dclab.new_dataset("data/example.rtdc")
-    kde = ds.get_kde_scatter(xax="area_um", yax="deform")
+    kde_instance = KernelDensityEstimator(ds)
+    kde = kde_instance.get_scatter(xax="area_um", yax="deform")
+    
     # load and apply polygon filter from file
     pf = dclab.PolygonFilter(filename="data/example.poly")
     ds.polygon_filter_add(pf)

{dclab-0.63.0 → dclab-0.63.1}/docs/sec_code_reference.rst

@@ -302,22 +302,18 @@ isoelastics
 
 
 
-.. _sec_ref_kde_contours:
+.. _sec_ref_kde:
 
-kde_contours
-------------
+kernel density estimators (KDEs)
+-------------------------------
+.. autoclass:: dclab.kde.KernelDensityEstimator
+   :members:
 
-.. automodule:: dclab.kde_contours
+.. automodule:: dclab.kde.methods
    :members:
    :undoc-members:
 
-
-.. _sec_ref_kde_methods:
-
-kde_methods
------------
-
-.. automodule:: dclab.kde_methods
+.. automodule:: dclab.kde.contours
    :members:
    :undoc-members: