uproot.AsStridedObjects

Defined in uproot.interpretation.objects on line 706.

Inheritance order:

  1. uproot.AsDtype

  2. uproot.interpretation.numerical.Numerical

  3. uproot.interpretation.Interpretation

class uproot.interpretation.objects.AsStridedObjects(model, members, original=None)
Parameters:

Interpretation for an array (possibly uproot.AsJagged) of fixed-size objects. Since the objects have a fixed number of fields with a fixed number of bytes each, the whole array (or content) can be interpreted in one vectorized array-cast. Therefore, this interpretation is faster than uproot.AsObjects when it is possible.

Unlike uproot.AsDtype with a structured array, the objects in the final array have the methods required by its model. If the library is uproot.interpretation.library.NumPy, these are instantiated as Python objects (slow); if uproot.interpretation.library.Awkward, they are behaviors passed to the Awkward Array’s local behavior.

model

AsStridedObjects.model

The full Uproot deserialization model for the data (uproot.Model or uproot.containers.AsContainer).

members

AsStridedObjects.members

The name (str) and fixed-width uproot.interpretation.Interpretation for each member of the objects as a list of 2-tuple pairs.

original

AsStridedObjects.original

If not None, this was the original model from an uproot.AsObjects that was simplified into this uproot.AsStridedObjects.

all_headers_prepended

AsStridedObjects.all_headers_prepended

from_dtype

Inherited from uproot.AsDtype.

AsStridedObjects.from_dtype

Data type to expect of the raw but uncompressed bytes in the TBasket data. Usually big-endian; may include named fields and a shape.

Named fields (dtype.names) can be used to construct a NumPy structured array.

A shape (dtype.shape) can be used to construct a fixed-size array for each entry. (Not applicable to variable-length lists! See uproot.AsJagged.) The finalized array’s array.shape[1:] == dtype.shape.

to_dtype

Inherited from uproot.AsDtype.

AsStridedObjects.to_dtype

Data type to convert the data into. Usually the native-endian equivalent of from_dtype; may include named fields and a shape.

Named fields (dtype.names) can be used to construct a NumPy structured array.

A shape (dtype.shape) can be used to construct a fixed-size array for each entry. (Not applicable to variable-length lists! See uproot.AsJagged.) The finalized array’s array.shape[1:] == dtype.shape.

itemsize

Inherited from uproot.AsDtype.

AsStridedObjects.itemsize

Number of bytes per item of from_dtype.

This number of bytes includes the fields and shape, like dtype.itemsize in NumPy.

inner_shape

Inherited from uproot.AsDtype.

AsStridedObjects.inner_shape

reshape

Inherited from uproot.AsDtype.

AsStridedObjects.reshape(shape)

inplace

Inherited from uproot.AsDtype.

AsStridedObjects.inplace(array)

Returns a AsDtypeInPlace version of self in order to fill the given array in place.

Example usage : ` var = np.zeros(N, dtype=np.float32) b = uproot.openn('afile.root')['treename']['varname'] b.array(library='np', interpretation=b.interpretation.inplace(var) ) `

cache_key

Inherited from uproot.interpretation.Interpretation.

AsStridedObjects.cache_key

typename

Inherited from uproot.interpretation.Interpretation.

AsStridedObjects.typename

numpy_dtype

Inherited from uproot.interpretation.Interpretation.

AsStridedObjects.numpy_dtype

awkward_form

Inherited from uproot.interpretation.Interpretation.

AsStridedObjects.awkward_form(file, context=None, index_format='i64', header=False, tobject_header=False, breadcrumbs=())
Parameters:
  • file (uproot.ReadOnlyFile) – File to use to generate uproot.Model classes from its streamers and file_path for error messages.

  • context (dict) – Context for the Form-generation; defaults are the remaining arguments below.

  • index_format (str) – Format to use for indexes of the awkward.forms.Form; may be "i32", "u32", or "i64".

  • header (bool) – If True, include header fields of each C++ class.

  • tobject_header (bool) – If True, include header fields of each TObject base class.

  • breadcrumbs (tuple of class objects) – Used to check for recursion. Types that contain themselves cannot be Awkward Arrays because the depth of instances is unknown.

The awkward.forms.Form to use to put objects of type type in an Awkward Array.

basket_array

Inherited from uproot.interpretation.Interpretation.

AsStridedObjects.basket_array(data, byte_offsets, basket, branch, context, cursor_offset, library, options)
Parameters:
  • data (array of numpy.uint8) – Raw but uncompressed data from the TBasket. If the TBasket has offsets and navigational metadata, it is not included in this array.

  • byte_offsets (array of numpy.int32) – Index where each entry of the TBasket starts and stops. The header is not included (i.e. the first offset is 0), and the length of this array is one greater than the number of entries in the TBasket.

  • basket (uproot.models.TBasket.Model_TBasket) – The TBasket object.

  • context (dict) – Auxiliary data used in deserialization.

  • cursor_offset (int) – Correction to the integer keys used in refs for objects deserialized by reference (uproot.deserialization.read_object_any).

  • library (uproot.interpretation.library.Library) – The requested library for output.

  • interp_options (dict) – Flags and other options passed through the interpretation process.

Performs the first step of interpretation, from uncompressed TBasket data to a temporary array.

final_array

Inherited from uproot.interpretation.Interpretation.

AsStridedObjects.final_array(basket_arrays, entry_start, entry_stop, entry_offsets, library, branch, options)
Parameters:
  • basket_arrays (dict of int → array) – Mapping from TBasket number to the temporary array returned by basket_array.

  • entry_start (int) – First entry to include when trimming any excess entries from the first TBasket.

  • entry_stop (int) – FIrst entry to exclude (one greater than the last entry to include) when trimming any excess entries from the last TBasket.

  • entry_offsets (list of int) – The entry_offsets for this TBranch.

  • library (uproot.interpretation.library.Library) – The requested library for output.

  • branch (uproot.TBranch) – The TBranch that is being interpreted.

  • interp_options (dict) – Flags and other options passed through the interpretation process.

Performs the last steps of interpretation, from a collection of temporary arrays, one for each TBasket, to a trimmed, finalized, grouped array, produced by the library.

hook_before_basket_array

Inherited from uproot.interpretation.Interpretation.

AsStridedObjects.hook_before_basket_array(*args, **kwargs)

Called in basket_array, before any interpretation.

This is the first hook called in basket_array.

hook_after_basket_array

Inherited from uproot.interpretation.Interpretation.

AsStridedObjects.hook_after_basket_array(*args, **kwargs)

Called in basket_array, after all interpretation.

This is the last hook called in basket_array.

hook_before_final_array

Inherited from uproot.interpretation.Interpretation.

AsStridedObjects.hook_before_final_array(*args, **kwargs)

Called in final_array, before any trimming, finalization, or grouping.

This is the first hook called in final_array.

hook_before_library_finalize

Inherited from uproot.interpretation.Interpretation.

AsStridedObjects.hook_before_library_finalize(*args, **kwargs)

Called in final_array, after trimming but before calling the finalize routine.

hook_after_final_array

Inherited from uproot.interpretation.Interpretation.

AsStridedObjects.hook_after_final_array(*args, **kwargs)

Called in final_array, after all trimming, finalization, and grouping.

This is the last hook called in final_array.