Differences

This shows you the differences between two versions of the page.

--- other:python:misc_by_jyp [2022/05/23 18:01] – [numpy related stuff] Changed the VIEW update warnng example jypeter
+++ other:python:misc_by_jyp [2022/07/08 16:00] – [numpy related stuff] Added the arbitrary object array jypeter
@@ Line 248: / Line 248: @@
 ==== numpy related stuff ====
+=== Using a numpy array to store arbitrary objects ===
+The numpy arrays are usually used to store [[https://numpy.org/doc/stable/reference/arrays.scalars.html|scalars]] of the same type (see also the [[https://numpy.org/doc/stable/reference/arrays.dtypes.html|Data type objects (dtype)]]), very often numerical values.
+It is also possible to store **arbitrary** Python objects in an array, rather than using nested lists or dictionaries!
+<code>>>> some_array = np.empty((2, 3), dtype=object)
+>>> some_array
+array([[None, None, None],
+       [None, None, None]], dtype=object)
+>>> some_array.shape
+(2, 3)
+>>> print(some_array[-1, -1])
+None
+>>> some_array[-1, 0] = filled_contour # e.g. save an existing cartopy filled contour object
+>>> some_array
+array([[None, None, None],
+       [<cartopy.mpl.contour.GeoContourSet object at 0x2ab679e8bf10>,
+        None, None]], dtype=object)</code>
 === Dealing with a variable number of indices ===