Skip to content

Commit

Permalink
Fixed examples Laehnemann using new schema
Browse files Browse the repository at this point in the history
  • Loading branch information
@atomprobe-tc
atomprobe-tc committed Sep 4, 2024
1 parent 8142fd2 commit cae676e
Show file tree
Hide file tree
Showing 4 changed files with 116 additions and 75 deletions.
14 changes: 10 additions & 4 deletions .vscode/launch.json
View file
Original file line number Diff line number Diff line change
Expand Up @@ -38,8 +38,11 @@
//"../apex/InGaN_nanowires_linescan.edaxh5",
//"../apex/InGaN_nanowires_map.edaxh5",
//"../apex/InGaN_nanowires_spectra.edaxh5",
//"tests/data/2023-08-16_Ni_NFDI.edaxh5",
"tests/data/hfive_oxford/173_0057.h5oina",
//"tests/data/hfive_apex/2023-08-16_Ni_NFDI.edaxh5",
//"tests/data/hfive_apex/InGaN_nanowires_map.edaxh5",
//"tests/data/hfive_apex/InGaN_nanowires_linescan.edaxh5",
"tests/data/hfive_apex/InGaN_nanowires_spectra.edaxh5",
//"tests/data/hfive_oxford/173_0057.h5oina",
"--reader",
"em",
"--nxdl",
Expand All @@ -57,8 +60,11 @@
//"--output=pynxtools_em/dbg/zeiss.nxs",
//"--output=pynxtools_em/dbg/tfs.nxs",
//"--output=dbg/apex.nxs"
"--output=dbg/hfive_oxford.nxs"]
//"--output=dbg/hfive_apex.nxs"],
//"--output=tests/data/hfive_apex/InGaN_nanowires_map.edaxh5.nxs"]
//"--output=tests/data/hfive_apex/InGaN_nanowires_linescan.edaxh5.nxs"]
"--output=tests/data/hfive_apex/InGaN_nanowires_spectra.edaxh5.nxs"]
//"--output=tests/data/hfive_apex/2023-08-16_Ni_NFDI.edaxh5.nxs"]
//"--output=tests/data/hfive_oxford/173_0057.h5oina.nxs"]
}
]
}
5 changes: 2 additions & 3 deletions src/pynxtools_em/methods/ebsd.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -501,8 +501,8 @@ def process_roi_phase_ipf(
template[f"{mpp}/@AXISNAME_indices[axis_{dim}_indices]"] = np.uint32(
dim_idx
)
template[f"{mpp}/DATA[data]"] = {"compress": ipf_rgb_map, "strength": 1}
hfive_web_decorate_nxdata(f"{mpp}/DATA[data]", template)
template[f"{mpp}/data"] = {"compress": ipf_rgb_map, "strength": 1}
hfive_web_decorate_nxdata(f"{mpp}/data", template)

# mind that EBSD map could be scale-invariant e.g. from synthetic microstructure
# simulation that could work on multiple length-scales as atoms are not resolved
Expand Down Expand Up @@ -533,7 +533,6 @@ def process_roi_phase_ipf(
dims = ["x", "y"] # no longer the EBSD map just an RGB image of the legend!
for dim in dims[::-1]:
template[f"{lgd}/@axes"].append(f"axis_{dim}")
enum = 0
for dim_idx, dim in enumerate(dims):
template[f"{lgd}/@AXISNAME_indices[axis_{dim}_indices]"] = np.uint32(
dim_idx
Expand Down
165 changes: 101 additions & 64 deletions src/pynxtools_em/parsers/hfive_apex.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -33,14 +33,21 @@
from pynxtools_em.utils.get_xrayline_iupac_names import get_xrayline_candidates
from pynxtools_em.utils.hfive_utils import read_strings
from pynxtools_em.utils.pint_custom_unit_registry import ureg
from ase.data import chemical_symbols


class HdfFiveEdaxApexParser(HdfFiveBaseParser):
"""Read APEX edaxh5"""

def __init__(self, file_path: str = "", entry_id: int = 1, verbose: bool = False):
super().__init__(file_path)
self.id_mgn: Dict[str, int] = {"entry_id": entry_id, "roi_id": 1}
self.id_mgn: Dict[str, int] = {
"entry_id": entry_id,
"event_id": 1,
"roi_id": 1,
"img_id": 1,
"spc_id": 1,
}
self.prfx: str = ""
self.tmp = {}
self.spc: Dict[str, Any] = {}
Expand Down Expand Up @@ -142,19 +149,21 @@ def parse(self, template: dict) -> dict:
if area_grp_nm.startswith("Live Map"):
# self.parse_and_normalize_eds_spd(h5r)
# element-specific ROI (aka element map)
self.parse_and_normalize_eds_area_rois(h5r)

if area_grp_nm.startswith(("LineScan", "ROILineScan")):
# "free form? or (which I assume) orthogonal line grid
# inside the FOV
# TODO::currently assume that internal organization of
# LineScan and ROILineScan groups is the same
# TODO but physical ROI they reference respectively differs
# TODO:: LineScan refers to the FOV that is
# in the parent of the LineScan group)
self.prfx = f"{abbrev}/{area_grp_nm}"
self.parse_and_normalize_eds_line_lsd(h5r)
# self.parse_and_normalize_eds_line_rois(h5r)
self.parse_and_normalize_eds_area_rois(
h5r, template
)

# if area_grp_nm.startswith(("LineScan", "ROILineScan")):
# "free form? or (which I assume) orthogonal line grid
# inside the FOV
# TODO::currently assume that internal organization of
# LineScan and ROILineScan groups is the same
# TODO but physical ROI they reference respectively differs
# TODO:: LineScan refers to the FOV that is
# in the parent of the LineScan group)
# self.prfx = f"{abbrev}/{area_grp_nm}"
# self.parse_and_normalize_eds_line_lsd(h5r)
# self.parse_and_normalize_eds_line_rois(h5r)
# TODO::make this nesting access code better readable although its benefit
# is that we do not need to visit first all nodes and do e.g. rfind operations
# full_hdf_path = "/a/b/c/d/Area" triggering action could also just be catched
Expand Down Expand Up @@ -418,8 +427,10 @@ def parse_and_normalize_eds_fov(self, fp, template: dict) -> dict:
),
}
# is micron because MicronsPerPixel{dim} used by EDAX
trg = f"/ENTRY[entry{self.id_mgn['entry_id']}]/measurement/event_data_em_set/EVENT_DATA_EM[event_data_em{self.id_mgn['event_id']}]/IMAGE_SET[image_set{self.id_mgn['img']}]"
template[f"{trg}/source"] = f"{self.prfx}/FOVIMAGE"
trg = f"/ENTRY[entry{self.id_mgn['entry_id']}]/measurement/event_data_em_set/EVENT_DATA_EM[event_data_em{self.id_mgn['event_id']}]/IMAGE_SET[image_set{self.id_mgn['img_id']}]"
template[f"{trg}/PROCESS[process]/source/absolute_path"] = (
f"{self.prfx}/FOVIMAGE"
)
template[f"{trg}/image_2d/@signal"] = "real"
template[f"{trg}/image_2d/@axes"] = ["axis_j", "axis_i"]
template[f"{trg}/image_2d/real"] = {
Expand All @@ -429,21 +440,25 @@ def parse_and_normalize_eds_fov(self, fp, template: dict) -> dict:
"strength": 1,
}
for dim_idx, dim in enumerate(["j", "i"]):
template[f"{trg}/image_2d/axis_{dim}"] = ureg.Quantity(
qnt = ureg.Quantity(
np.asarray(
np.linspace(0, nyx[dim] - 1, num=nyx[dim], endpoint=True)
* syx[dim].magnitude,
dtype=syx[dim].dtype,
),
syx[dim].units,
)
template[f"{trg}/image_2d/axis_{dim}"] = {
"compress": qnt.magnitude,
"strength": 1,
}
template[f"{trg}/image_2d/axis_{dim}/@long_name"] = (
f"### Position along {dim} ({syx[dim].units})"
f"Point coordinate along axis-{dim} ({qnt.units})"
)
template[f"{trg}/image_2d/@AXISNAME_indices[axis_{dim}_indices]"] = (
np.uint32(dim_idx)
)
self.id_mgn["img"] += 1
self.id_mgn["img_id"] += 1
return template

def parse_and_normalize_eds_spc(self, fp, template: dict) -> dict:
Expand All @@ -462,11 +477,11 @@ def parse_and_normalize_eds_spc(self, fp, template: dict) -> dict:
print(f"Attribute {req} not found in {self.prfx}/SPC !")
return template

trg = f"/ENTRY[entry{self.id_mgn['entry_id']}]/measurement/event_data_em_set/EVENT_DATA_EM[event_data_em{self.id_mgn['event_id']}]/SPECTRUM_SET[spectrum_set{self.id_mgn['img']}]"
template[f"{trg}/source"] = f"{self.prfx}/SPC"
trg = f"/ENTRY[entry{self.id_mgn['entry_id']}]/measurement/event_data_em_set/EVENT_DATA_EM[event_data_em{self.id_mgn['event_id']}]/SPECTRUM_SET[spectrum_set{self.id_mgn['spc_id']}]"
template[f"{trg}/PROCESS[process]/source/absolute_path"] = f"{self.prfx}/SPC"
template[f"{trg}/spectrum_0d/@signal"] = "intensity"
template[f"{trg}/spectrum_0d/@axes"] = ["axis_energy"]
template[f"{trg}/spectrum_0d/@AXISNAME_indices[axis_energy_indices"] = (
template[f"{trg}/spectrum_0d/@AXISNAME_indices[axis_energy_indices]"] = (
np.uint32(0)
)
e_zero = ureg.Quantity(fp[f"{self.prfx}/SPC"]["eVOffset"][0], ureg.eV)
Expand All @@ -485,52 +500,52 @@ def parse_and_normalize_eds_spc(self, fp, template: dict) -> dict:
.to(ureg.keV)
.magnitude
)
template[f"{trg}/spectrum_0d/axis_energy/@long_name"] = (
f"Energy ({ureg.keV.units})"
)
template[f"{trg}/spectrum_0d/axis_energy/@long_name"] = f"Energy ({ureg.keV})"
template[f"{trg}/spectrum_0d/intensity"] = {
"compress": np.asarray(
fp[f"{self.prfx}/SPC"]["SpectrumCounts"][0], dtype=np.int32
),
"strength": 1,
}
template[f"{trg}/spectrum_0d/intensity/@long_name"] = f"Counts"
self.id_mgn["spc_id"] += 1
return template

def parse_and_normalize_eds_area_rois(self, fp):
def parse_and_normalize_eds_area_rois(self, fp, template: dict) -> dict:
"""Normalize and scale APEX-specific EDS element line maps to NeXus."""
print(f"Parsing {self.prfx} ...")
for req in ["ELEMENTOVRLAYIMGCOLLECTIONPARAMS", "PHASES", "ROIs", "SPC"]:
if f"{self.prfx}/{req}" not in fp:
print(f"Group {self.prfx}/{req} not found !")
return
return template
for req in ["eVOffset", "evPch", "NumberOfPoints", "SpectrumCounts"]:
if req not in fp[f"{self.prfx}/SPC"].dtype.names: # also check for shape
print(f"Attribute {req} not found in {self.prfx}/SPC !")
return
return template
for req in ["ResolutionX", "ResolutionY", "mmFieldWidth", "mmFieldHeight"]:
if (
req
not in fp[f"{self.prfx}/ELEMENTOVRLAYIMGCOLLECTIONPARAMS"].dtype.names
):
abbrev = f"{self.prfx}/ELEMENTOVRLAYIMGCOLLECTIONPARAMS"
if req not in fp[abbrev].dtype.names:
# also check for shape
print(
f"Attribute {req} not found in {self.prfx}/ELEMENTOVRLAYIMGCOLLECTIONPARAMS !"
)
# find relevant EDS maps (pairs of <symbol>.dat, <symbol>.ipr) groups
uniq = set()
print(f"Attribute {req} not found in {abbrev} !")
return template
# find relevant EDS maps via demanding at least one pair of
# <symbol>.dat, <symbol>.ipr HDF5 group respectively
pairs = set()
for group_name in fp[f"{self.prfx}/ROIs"]:
token = group_name.split(".")
if len(token) == 2 and token[1] in ("dat", "ipr"):
uniq.add(token[0])
for entry in uniq:
if len(token) == 2 and token[1] in ("dat", "ipr") and token[0] not in pairs:
pairs.add(token[0])
for pair in pairs:
if (
(f"{self.prfx}/ROIs/{entry}.dat" not in fp[f"{self.prfx}/ROIs"])
or (f"{self.prfx}/ROIs/{entry}.ipr" not in fp[f"{self.prfx}/ROIs"])
or ("RoiStartChan" not in fp[f"{self.prfx}/ROIs/{entry}.dat"].attrs)
or ("RoiEndChan" not in fp[f"{self.prfx}/ROIs/{entry}.dat"].attrs)
(f"{self.prfx}/ROIs/{pair}.dat" not in fp[f"{self.prfx}/ROIs"])
or (f"{self.prfx}/ROIs/{pair}.ipr" not in fp[f"{self.prfx}/ROIs"])
or ("RoiStartChan" not in fp[f"{self.prfx}/ROIs/{pair}.dat"].attrs)
or ("RoiEndChan" not in fp[f"{self.prfx}/ROIs/{pair}.dat"].attrs)
or pair[0 : pair.find(" ")] not in chemical_symbols[1:]
):
uniq.remove(entry)
pairs.remove(pair)
if len(pairs) == 0:
return template

e_zero = ureg.Quantity(fp[f"{self.prfx}/SPC"]["eVOffset"][0], ureg.eV)
e_delta = ureg.Quantity(fp[f"{self.prfx}/SPC"]["evPch"][0], ureg.eV)
Expand All @@ -543,7 +558,7 @@ def parse_and_normalize_eds_area_rois(self, fp):
dtype=e_zero.magnitude.dtype,
),
ureg.eV,
) # eV, as xraydb demands
) # eV, as xraydb demands eV for finding line candidates
nxy = {
"i": fp[f"{self.prfx}/ELEMENTOVRLAYIMGCOLLECTIONPARAMS"][0]["ResolutionX"],
"j": fp[f"{self.prfx}/ELEMENTOVRLAYIMGCOLLECTIONPARAMS"][0]["ResolutionY"],
Expand All @@ -558,26 +573,38 @@ def parse_and_normalize_eds_area_rois(self, fp):
"i": ureg.Quantity(nxy["li"] / nxy["i"], ureg.millimeter),
"j": ureg.Quantity(nxy["lj"] / nxy["j"], ureg.millimeter),
}
self.eds = []
for entry in uniq:
eds_map = {}
eds_map["source"] = f"{self.prfx}/ROIs/{entry}"
eds_map["description"] = f"{entry}"
for pair in pairs:
trg = f"/ENTRY[entry{self.id_mgn['entry_id']}]/roiID[roi{self.id_mgn['roi_id']}]/eds/indexing/IMAGE_SET[{pair[0:pair.find(' ')]}]"
template[f"{trg}/PROCESS[process]/source/absolute_path"] = (
f"{self.prfx}/ROIs/{pair}"
)
# this can be a custom name e.g. InL or In L but it is not necessarily
# a clean description of an element plus a IUPAC line, hence get all
# theoretical candidates within integrated energy region [e_roi_s, e_roi_e]
e_roi_s = fp[f"{self.prfx}/ROIs/{entry}.dat"].attrs["RoiStartChan"][0]
e_roi_e = fp[f"{self.prfx}/ROIs/{entry}.dat"].attrs["RoiEndChan"][0]
eds_map["energy_range"] = ureg.Quantity(
np.asarray([e_channels[e_roi_s], e_channels[e_roi_e + 1]]), ureg.eV
e_roi_s = fp[f"{self.prfx}/ROIs/{pair}.dat"].attrs["RoiStartChan"][0]
e_roi_e = fp[f"{self.prfx}/ROIs/{pair}.dat"].attrs["RoiEndChan"][0]
rng = ureg.Quantity(
np.asarray(
[e_channels.magnitude[e_roi_s], e_channels.magnitude[e_roi_e + 1]]
),
ureg.eV,
)
eds_map.tmp["iupac_line_candidates"] = ", ".join(
template[f"{trg}/energy_range"] = rng.magnitude
template[f"{trg}/energy_range/@units"] = f"{rng.units}"
template[f"{trg}/iupac_line_candidates"] = ", ".join(
get_xrayline_candidates(
e_channels.magnitude[e_roi_s], e_channels.magnitude[e_roi_e + 1]
)
)
for dim in ["i", "j"]:
eds_map[f"image_2d/axis_{dim}"] = ureg.Quantity(
template[f"{trg}/image_2d/@signal"] = "intensity"
template[f"{trg}/image_2d/@axes"] = ["axis_j", "axis_i"]
template[f"{trg}/image_2d/title"] = f"{pair}"
template[f"{trg}/image_2d/intensity"] = {
"compress": np.asarray(fp[f"{self.prfx}/ROIs/{pair}.dat"]),
"strength": 1,
}
for dim_idx, dim in enumerate(["i", "j"]):
qnt = ureg.Quantity(
np.asarray(
0.0
+ np.linspace(0, nxy[dim] - 1, num=int(nxy[dim]), endpoint=True)
Expand All @@ -586,14 +613,23 @@ def parse_and_normalize_eds_area_rois(self, fp):
),
sxy[dim].units,
)
eds_map[f"image_2d/axis_{dim}/@long_name"] = (
f"Point coordinate along the {dim}-axis ({sxy[dim].units})"
template[f"{trg}/image_2d/axis_{dim}"] = {
"compress": qnt.magnitude,
"strength": 1,
}
template[f"{trg}/image_2d/@AXISNAME_indices[axis_{dim}_indices]"] = (
np.uint32(dim_idx)
)
eds_map["image_2d/real"] = np.asarray(
fp[f"{self.prfx}/ROIs/{entry}.dat"]
template[f"{trg}/image_2d/axis_{dim}/@long_name"] = (
f"Point coordinate along the {dim}-axis ({qnt.units})"
)
self.eds.append(eds_map)
template[f"{trg}/image_2d/axis_{dim}/@units"] = f"{qnt.units}"
return template

# TODO::these functions were deactivated as they have few examples and have not been
# discussed properly, below code is rather meant as snippets useful for a future
# extension
'''
def parse_and_normalize_eds_line_lsd(self, fp):
"""Normalize and scale APEX-specific line scan with one spectrum each to NeXus."""
# https://hyperspy.org/rosettasciio/_downloads/
Expand Down Expand Up @@ -680,7 +716,7 @@ def parse_and_normalize_eds_spd(self, fp):
# c2e8b23d511a3c44fc30c69114e2873e/SpcMap-spd.file.format.pdf
print(f"Parsing {self.prfx} ...")
if f"{self.prfx}/SPD" not in fp:
print(f"Group {self.prfx/SPD} not found !")
print(f"Group {self.prfx}/SPD not found !")
return
for req in [
Expand Down Expand Up @@ -756,3 +792,4 @@ def parse_and_normalize_eds_spd(self, fp):
# pure speculation and because of the fact that not even the SPD file format
# specification details the metadata, i.e. energy per channel, start and end
# we do not use the SPD instance right now
'''
7 changes: 3 additions & 4 deletions src/pynxtools_em/utils/get_xrayline_iupac_names.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@
import xraydb

from ase.data import chemical_symbols
from typing import List


def get_all_xraylines() -> dict:
Expand All @@ -30,14 +31,12 @@ def get_all_xraylines() -> dict:
return xray_lines


def get_xrayline_candidates(e_min, e_max) -> list:
def get_xrayline_candidates(e_min, e_max) -> List[str]:
# one could try to resolve the line from the alias of
# the actual entry but this is not rigorous!
cand = []
for key, val in get_all_xraylines().items():
if val < e_min:
continue
if val > e_max:
if val < e_min or val > e_max:
continue
cand.append(key)
return cand

0 comments on commit cae676e

Please sign in to comment.