def hit2df(
fname,
cache=True,
verbose=True,
drop_non_numeric=True,
load_wavenum_min=None,
load_wavenum_max=None,
):
"""Convert a HITRAN/HITEMP [1]_ file to a Pandas dataframe
Parameters
----------
fname: str
HITRAN-HITEMP file name
cache: boolean, or ``'regen'`` or ``'force'``
if ``True``, a pandas-readable HDF5 file is generated on first access,
and later used. This saves on the datatype cast and conversion and
improves performances a lot (but changes in the database are not
taken into account). If False, no database is used. If ``'regen'``, temp
file are reconstructed. Default ``True``.
Other Parameters
----------------
drop_non_numeric: boolean
if ``True``, non numeric columns are dropped. This improves performances,
but make sure all the columns you need are converted to numeric formats
before hand. Default ``True``. Note that if a cache file is loaded it
will be left untouched.
load_wavenum_min, load_wavenum_max: float
if not ``'None'``, only load the cached file if it contains data for
wavenumbers above/below the specified value. See :py:func`~radis.io.cache_files.load_h5_cache_file`.
Default ``'None'``.
Returns
-------
df: pandas Dataframe
dataframe containing all lines and parameters
References
----------
.. [1] `HITRAN 1996, Rothman et al., 1998 <https://www.sciencedirect.com/science/article/pii/S0022407398000788>`__
Notes
-----
Performances: see CDSD-HITEMP parser
See Also
--------
:func:`~radis.io.cdsd.cdsd2df`
"""
metadata = {}
# Last modification time of the original file :
metadata["last_modification"] = time.ctime(getmtime(fname))
if verbose >= 2:
print("Opening file {0} (cache={1})".format(fname, cache))
print("Last modification time: {0}".format(metadata["last_modification"]))
if load_wavenum_min and load_wavenum_max:
assert load_wavenum_min < load_wavenum_max
columns = columns_2004
# Use cache file if possible
fcache = cache_file_name(fname)
if cache and exists(fcache):
relevant_if_metadata_above = (
{"wavenum_max": load_wavenum_min} if load_wavenum_min else {}
) # not relevant if wavenum_max of file is < wavenum min required
relevant_if_metadata_below = (
{"wavenum_min": load_wavenum_max} if load_wavenum_max else {}
) # not relevant if wavenum_min of file is > wavenum max required
df = load_h5_cache_file(
fcache,
cache,
valid_if_metadata_is=metadata,
relevant_if_metadata_above=relevant_if_metadata_above,
relevant_if_metadata_below=relevant_if_metadata_below,
current_version=radis.__version__,
last_compatible_version=OLDEST_COMPATIBLE_VERSION,
verbose=verbose,
)
if df is not None:
return df
# Detect the molecule by reading the start of the file
try:
with open(fname) as f:
mol = get_molecule(int(f.read(2)))
except UnicodeDecodeError as err:
raise ValueError(
"You're trying to read a binary file {0} ".format(fname)
+ "instead of an HITRAN file"
) from err
# %% Start reading the full file
df = parse_hitran_file(fname, columns)
# %% Post processing
# assert one molecule per database only. Else the groupbase data reading
# above doesnt make sense
nmol = len(set(df["id"]))
if nmol == 0:
raise ValueError("Databank looks empty")
elif nmol != 1:
# Crash, give explicity error messages
try:
secondline = df.iloc[1]
except IndexError:
secondline = ""
raise ValueError(
"Multiple molecules in database ({0}). Current ".format(nmol)
+ "spectral code only computes 1 species at the time. Use MergeSlabs. "
+ "Verify the parsing was correct by looking at the first row below: "
+ "\n{0}".format(df.iloc[0])
+ "\n----------------\nand the second row "
+ "below: \n{0}".format(secondline)
)
# Add local quanta attributes, based on the HITRAN group
df = parse_local_quanta(df, mol)
# Add global quanta attributes, based on the HITRAN class
df = parse_global_quanta(df, mol)
# Remove non numerical attributes
if drop_non_numeric:
if "branch" in df:
replace_PQR_with_m101(df)
df = drop_object_format_columns(df, verbose=verbose)
# cached file mode but cached file doesn't exist yet (else we had returned)
if cache:
new_metadata = {
# Last modification time of the original file :
"last_modification": time.ctime(getmtime(fname)),
"wavenum_min": df.wav.min(),
"wavenum_max": df.wav.max(),
}
if verbose:
print(
"Generating cache file {0} with metadata :\n{1}".format(
fcache, new_metadata
)
)
try:
save_to_hdf(
df,
fcache,
metadata=new_metadata,
version=radis.__version__,
key="df",
overwrite=True,
verbose=verbose,
)
except PermissionError:
if verbose:
print(sys.exc_info())
print("An error occured in cache file generation. Lookup access rights")
pass
# TODO : get only wavenum above/below 'load_wavenum_min', 'load_wavenum_max'
# by parsing df.wav. Completely irrelevant files are discarded in 'load_h5_cache_file'
# but files that have partly relevant lines are fully loaded.
# Note : cache file is generated with the full line list.
return df
def hit2df(fname, count=-1, cache=False, verbose=True, drop_non_numeric=True):
""" Convert a HITRAN/HITEMP [1]_ file to a Pandas dataframe
Parameters
----------
fname: str
HITRAN-HITEMP file name
count: int
number of items to read (-1 means all file)
cache: boolean, or ``'regen'`` or ``'force'``
if ``True``, a pandas-readable HDF5 file is generated on first access,
and later used. This saves on the datatype cast and conversion and
improves performances a lot (but changes in the database are not
taken into account). If False, no database is used. If ``'regen'``, temp
file are reconstructed. Default ``False``.
Other Parameters
----------------
drop_non_numeric: boolean
if ``True``, non numeric columns are dropped. This improves performances,
but make sure all the columns you need are converted to numeric formats
before hand. Default ``True``. Note that if a cache file is loaded it
will be left untouched.
Returns
-------
df: pandas Dataframe
dataframe containing all lines and parameters
References
----------
.. [1] `HITRAN 1996, Rothman et al., 1998 <https://www.sciencedirect.com/science/article/pii/S0022407398000788>`__
Notes
-----
Performances: see CDSD-HITEMP parser
See Also
--------
:func:`~radis.io.cdsd.cdsd2df`
"""
if verbose >= 2:
print("Opening file {0} (cache={1})".format(fname, cache))
columns = columns_2004
# Use cache file if possible
fcache = splitext(fname)[0] + ".h5"
check_cache_file(fcache=fcache, use_cached=cache, verbose=verbose)
if cache and exists(fcache):
return get_cache_file(fcache, verbose=verbose)
# Detect the molecule by reading the start of the file
try:
with open(fname) as f:
mol = get_molecule(int(f.read(2)))
except UnicodeDecodeError as err:
raise ValueError(
"You're trying to read a binary file {0} ".format(fname) +
"instead of an HITRAN file") from err
# %% Start reading the full file
df = parse_hitran_file(fname, columns, count)
# %% Post processing
# assert one molecule per database only. Else the groupbase data reading
# above doesnt make sense
nmol = len(set(df["id"]))
if nmol == 0:
raise ValueError("Databank looks empty")
elif nmol != 1:
# Crash, give explicity error messages
try:
secondline = df.iloc[1]
except IndexError:
secondline = ""
raise ValueError(
"Multiple molecules in database ({0}). Current ".format(nmol) +
"spectral code only computes 1 species at the time. Use MergeSlabs. "
+
"Verify the parsing was correct by looking at the first row below: "
+ "\n{0}".format(df.iloc[0]) +
"\n----------------\nand the second row " +
"below: \n{0}".format(secondline))
# dd local quanta attributes, based on the HITRAN group
df = parse_local_quanta(df, mol)
# Add global quanta attributes, based on the HITRAN class
df = parse_global_quanta(df, mol)
# Remove non numerical attributes
if drop_non_numeric:
if "branch" in df:
replace_PQR_with_m101(df)
df = drop_object_format_columns(df, verbose=verbose)
# cached file mode but cached file doesn't exist yet (else we had returned)
if cache:
if verbose:
print("Generating cached file: {0}".format(fcache))
try:
save_to_hdf(
df,
fcache,
metadata={},
version=radis.__version__,
key="df",
overwrite=True,
verbose=verbose,
)
except:
if verbose:
print(sys.exc_info())
print(
"An error occured in cache file generation. Lookup access rights"
)
pass
return df
def cdsd2df(fname,
version="hitemp",
count=-1,
cache=False,
verbose=True,
drop_non_numeric=True):
"""Convert a CDSD-HITEMP [1]_ or CDSD-4000 [2]_ file to a Pandas dataframe
Parameters
----------
fname: str
CDSD file name
version: str ('4000', 'hitemp')
CDSD version
count: int
number of items to read (-1 means all file)
cache: boolean, or 'regen'
if ``True``, a pandas-readable HDF5 file is generated on first access,
and later used. This saves on the datatype cast and conversion and
improves performances a lot (but changes in the database are not
taken into account). If ``False``, no database is used. If 'regen', temp
file are reconstructed. Default ``False``.
Other Parameters
----------------
drop_non_numeric: boolean
if ``True``, non numeric columns are dropped. This improves performances,
but make sure all the columns you need are converted to numeric formats
before hand. Default ``True``. Note that if a cache file is loaded it
will be left untouched.
Returns
-------
df: pandas Dataframe
dataframe containing all lines and parameters
Notes
-----
CDSD-4000 Database can be downloaded from [3]_
Performances: I had huge performance trouble with this function, because the files are
huge (500k lines) and the format is to special (no space between numbers...)
to apply optimized methods such as pandas's. A line by line reading isn't
so bad, using struct to parse each line. However, we waste typing determining
what every line is. I ended up using the fromfiles functions from numpy,
not considering *\\n* (line return) as a special character anymore, and a second call
to numpy to cast the correct format. That ended up being twice as fast.
- initial: 20s / loop
- with mmap: worse
- w/o readline().rstrip('\\n'): still 20s
- numpy fromfiles: 17s
- no more readline, 2x fromfile 9s
Think about using cache mode too:
- no cache mode 9s
- cache mode, first time 22s
- cache mode, then 2s
Moving to HDF5:
On cdsd_02069_02070 (56 Mb)
Reading::
cdsd2df(): 9.29 s
cdsd2df(cache=True [old .txt version]): 2.3s
cdsd2df(cache=True [new h5 version, table]): 910ms
cdsd2df(cache=True [new h5 version, fixed]): 125ms
Storage::
%timeit df.to_hdf("cdsd_02069_02070.h5", "df", format="fixed") 337ms
%timeit df.to_hdf("cdsd_02069_02070.h5", "df", format="table") 1.03s
References
----------
Note that CDSD-HITEMP is used as the line database for CO2 in HITEMP 2010
.. [1] `HITEMP 2010, Rothman et al., 2010 <https://www.sciencedirect.com/science/article/pii/S002240731000169X>`_
.. [2] `CDSD-4000 article, Tashkun et al., 2011 <https://www.sciencedirect.com/science/article/pii/S0022407311001154>`_
.. [3] `CDSD-4000 database <ftp://ftp.iao.ru/pub/CDSD-4000/>`_
See Also
--------
:func:`~radis.io.hitran.hit2df`
"""
metadata = {}
metadata["last_modification"] = time.ctime(getmtime(fname))
if verbose >= 2:
print("Opening file {0} (format=CDSD {1}, cache={2})".format(
fname, version, cache))
print("Last Modification time: {0}".format(
metadata["last_modification"]))
if version == "hitemp":
columns = columns_hitemp
elif version == "4000":
columns = columns_4000
else:
raise ValueError("Unknown CDSD version: {0}".format(version))
# Use cache file if possible
fcache = cache_file_name(fname)
if cache and exists(fcache):
df = load_h5_cache_file(
fcache,
cache,
metadata=metadata,
current_version=radis.__version__,
last_compatible_version=OLDEST_COMPATIBLE_VERSION,
verbose=verbose,
)
if df is not None:
return df
# %% Start reading the full file
df = parse_hitran_file(fname, columns, count)
# Remove non numerical attributes
if drop_non_numeric:
replace_PQR_with_m101(df)
df = drop_object_format_columns(df, verbose=verbose)
# cached file mode but cached file doesn't exist yet (else we had returned)
if cache:
if verbose:
print("Generating cached file: {0}".format(fcache))
try:
save_to_hdf(
df,
fcache,
metadata=metadata,
version=radis.__version__,
key="df",
overwrite=True,
verbose=verbose,
)
except:
if verbose:
print(
"An error occured in cache file generation. Lookup access rights"
)
pass
return df