File size: 25,752 Bytes
7885a28 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 |
# Last Change: Mon Aug 20 08:00 PM 2007 J
import re
import datetime
import numpy as np
import csv
import ctypes
"""A module to read arff files."""
__all__ = ['MetaData', 'loadarff', 'ArffError', 'ParseArffError']
# An Arff file is basically two parts:
# - header
# - data
#
# A header has each of its components starting by @META where META is one of
# the keyword (attribute of relation, for now).
# TODO:
# - both integer and reals are treated as numeric -> the integer info
# is lost!
# - Replace ValueError by ParseError or something
# We know can handle the following:
# - numeric and nominal attributes
# - missing values for numeric attributes
r_meta = re.compile(r'^\s*@')
# Match a comment
r_comment = re.compile(r'^%')
# Match an empty line
r_empty = re.compile(r'^\s+$')
# Match a header line, that is a line which starts by @ + a word
r_headerline = re.compile(r'^\s*@\S*')
r_datameta = re.compile(r'^@[Dd][Aa][Tt][Aa]')
r_relation = re.compile(r'^@[Rr][Ee][Ll][Aa][Tt][Ii][Oo][Nn]\s*(\S*)')
r_attribute = re.compile(r'^\s*@[Aa][Tt][Tt][Rr][Ii][Bb][Uu][Tt][Ee]\s*(..*$)')
r_nominal = re.compile(r'{(.+)}')
r_date = re.compile(r"[Dd][Aa][Tt][Ee]\s+[\"']?(.+?)[\"']?$")
# To get attributes name enclosed with ''
r_comattrval = re.compile(r"'(..+)'\s+(..+$)")
# To get normal attributes
r_wcomattrval = re.compile(r"(\S+)\s+(..+$)")
# ------------------------
# Module defined exception
# ------------------------
class ArffError(OSError):
pass
class ParseArffError(ArffError):
pass
# ----------
# Attributes
# ----------
class Attribute:
type_name = None
def __init__(self, name):
self.name = name
self.range = None
self.dtype = np.object_
@classmethod
def parse_attribute(cls, name, attr_string):
"""
Parse the attribute line if it knows how. Returns the parsed
attribute, or None.
"""
return None
def parse_data(self, data_str):
"""
Parse a value of this type.
"""
return None
def __str__(self):
"""
Parse a value of this type.
"""
return self.name + ',' + self.type_name
class NominalAttribute(Attribute):
type_name = 'nominal'
def __init__(self, name, values):
super().__init__(name)
self.values = values
self.range = values
self.dtype = (np.bytes_, max(len(i) for i in values))
@staticmethod
def _get_nom_val(atrv):
"""Given a string containing a nominal type, returns a tuple of the
possible values.
A nominal type is defined as something framed between braces ({}).
Parameters
----------
atrv : str
Nominal type definition
Returns
-------
poss_vals : tuple
possible values
Examples
--------
>>> from scipy.io.arff._arffread import NominalAttribute
>>> NominalAttribute._get_nom_val("{floup, bouga, fl, ratata}")
('floup', 'bouga', 'fl', 'ratata')
"""
m = r_nominal.match(atrv)
if m:
attrs, _ = split_data_line(m.group(1))
return tuple(attrs)
else:
raise ValueError("This does not look like a nominal string")
@classmethod
def parse_attribute(cls, name, attr_string):
"""
Parse the attribute line if it knows how. Returns the parsed
attribute, or None.
For nominal attributes, the attribute string would be like '{<attr_1>,
<attr2>, <attr_3>}'.
"""
if attr_string[0] == '{':
values = cls._get_nom_val(attr_string)
return cls(name, values)
else:
return None
def parse_data(self, data_str):
"""
Parse a value of this type.
"""
if data_str in self.values:
return data_str
elif data_str == '?':
return data_str
else:
raise ValueError(f"{str(data_str)} value not in {str(self.values)}")
def __str__(self):
msg = self.name + ",{"
for i in range(len(self.values)-1):
msg += self.values[i] + ","
msg += self.values[-1]
msg += "}"
return msg
class NumericAttribute(Attribute):
def __init__(self, name):
super().__init__(name)
self.type_name = 'numeric'
self.dtype = np.float64
@classmethod
def parse_attribute(cls, name, attr_string):
"""
Parse the attribute line if it knows how. Returns the parsed
attribute, or None.
For numeric attributes, the attribute string would be like
'numeric' or 'int' or 'real'.
"""
attr_string = attr_string.lower().strip()
if (attr_string[:len('numeric')] == 'numeric' or
attr_string[:len('int')] == 'int' or
attr_string[:len('real')] == 'real'):
return cls(name)
else:
return None
def parse_data(self, data_str):
"""
Parse a value of this type.
Parameters
----------
data_str : str
string to convert
Returns
-------
f : float
where float can be nan
Examples
--------
>>> from scipy.io.arff._arffread import NumericAttribute
>>> atr = NumericAttribute('atr')
>>> atr.parse_data('1')
1.0
>>> atr.parse_data('1\\n')
1.0
>>> atr.parse_data('?\\n')
nan
"""
if '?' in data_str:
return np.nan
else:
return float(data_str)
def _basic_stats(self, data):
nbfac = data.size * 1. / (data.size - 1)
return (np.nanmin(data), np.nanmax(data),
np.mean(data), np.std(data) * nbfac)
class StringAttribute(Attribute):
def __init__(self, name):
super().__init__(name)
self.type_name = 'string'
@classmethod
def parse_attribute(cls, name, attr_string):
"""
Parse the attribute line if it knows how. Returns the parsed
attribute, or None.
For string attributes, the attribute string would be like
'string'.
"""
attr_string = attr_string.lower().strip()
if attr_string[:len('string')] == 'string':
return cls(name)
else:
return None
class DateAttribute(Attribute):
def __init__(self, name, date_format, datetime_unit):
super().__init__(name)
self.date_format = date_format
self.datetime_unit = datetime_unit
self.type_name = 'date'
self.range = date_format
self.dtype = np.datetime64(0, self.datetime_unit)
@staticmethod
def _get_date_format(atrv):
m = r_date.match(atrv)
if m:
pattern = m.group(1).strip()
# convert time pattern from Java's SimpleDateFormat to C's format
datetime_unit = None
if "yyyy" in pattern:
pattern = pattern.replace("yyyy", "%Y")
datetime_unit = "Y"
elif "yy":
pattern = pattern.replace("yy", "%y")
datetime_unit = "Y"
if "MM" in pattern:
pattern = pattern.replace("MM", "%m")
datetime_unit = "M"
if "dd" in pattern:
pattern = pattern.replace("dd", "%d")
datetime_unit = "D"
if "HH" in pattern:
pattern = pattern.replace("HH", "%H")
datetime_unit = "h"
if "mm" in pattern:
pattern = pattern.replace("mm", "%M")
datetime_unit = "m"
if "ss" in pattern:
pattern = pattern.replace("ss", "%S")
datetime_unit = "s"
if "z" in pattern or "Z" in pattern:
raise ValueError("Date type attributes with time zone not "
"supported, yet")
if datetime_unit is None:
raise ValueError("Invalid or unsupported date format")
return pattern, datetime_unit
else:
raise ValueError("Invalid or no date format")
@classmethod
def parse_attribute(cls, name, attr_string):
"""
Parse the attribute line if it knows how. Returns the parsed
attribute, or None.
For date attributes, the attribute string would be like
'date <format>'.
"""
attr_string_lower = attr_string.lower().strip()
if attr_string_lower[:len('date')] == 'date':
date_format, datetime_unit = cls._get_date_format(attr_string)
return cls(name, date_format, datetime_unit)
else:
return None
def parse_data(self, data_str):
"""
Parse a value of this type.
"""
date_str = data_str.strip().strip("'").strip('"')
if date_str == '?':
return np.datetime64('NaT', self.datetime_unit)
else:
dt = datetime.datetime.strptime(date_str, self.date_format)
return np.datetime64(dt).astype(
f"datetime64[{self.datetime_unit}]")
def __str__(self):
return super().__str__() + ',' + self.date_format
class RelationalAttribute(Attribute):
def __init__(self, name):
super().__init__(name)
self.type_name = 'relational'
self.dtype = np.object_
self.attributes = []
self.dialect = None
@classmethod
def parse_attribute(cls, name, attr_string):
"""
Parse the attribute line if it knows how. Returns the parsed
attribute, or None.
For date attributes, the attribute string would be like
'date <format>'.
"""
attr_string_lower = attr_string.lower().strip()
if attr_string_lower[:len('relational')] == 'relational':
return cls(name)
else:
return None
def parse_data(self, data_str):
# Copy-pasted
elems = list(range(len(self.attributes)))
escaped_string = data_str.encode().decode("unicode-escape")
row_tuples = []
for raw in escaped_string.split("\n"):
row, self.dialect = split_data_line(raw, self.dialect)
row_tuples.append(tuple(
[self.attributes[i].parse_data(row[i]) for i in elems]))
return np.array(row_tuples,
[(a.name, a.dtype) for a in self.attributes])
def __str__(self):
return (super().__str__() + '\n\t' +
'\n\t'.join(str(a) for a in self.attributes))
# -----------------
# Various utilities
# -----------------
def to_attribute(name, attr_string):
attr_classes = (NominalAttribute, NumericAttribute, DateAttribute,
StringAttribute, RelationalAttribute)
for cls in attr_classes:
attr = cls.parse_attribute(name, attr_string)
if attr is not None:
return attr
raise ParseArffError(f"unknown attribute {attr_string}")
def csv_sniffer_has_bug_last_field():
"""
Checks if the bug https://bugs.python.org/issue30157 is unpatched.
"""
# We only compute this once.
has_bug = getattr(csv_sniffer_has_bug_last_field, "has_bug", None)
if has_bug is None:
dialect = csv.Sniffer().sniff("3, 'a'")
csv_sniffer_has_bug_last_field.has_bug = dialect.quotechar != "'"
has_bug = csv_sniffer_has_bug_last_field.has_bug
return has_bug
def workaround_csv_sniffer_bug_last_field(sniff_line, dialect, delimiters):
"""
Workaround for the bug https://bugs.python.org/issue30157 if is unpatched.
"""
if csv_sniffer_has_bug_last_field():
# Reuses code from the csv module
right_regex = r'(?P<delim>[^\w\n"\'])(?P<space> ?)(?P<quote>["\']).*?(?P=quote)(?:$|\n)' # noqa: E501
for restr in (r'(?P<delim>[^\w\n"\'])(?P<space> ?)(?P<quote>["\']).*?(?P=quote)(?P=delim)', # ,".*?", # noqa: E501
r'(?:^|\n)(?P<quote>["\']).*?(?P=quote)(?P<delim>[^\w\n"\'])(?P<space> ?)', # .*?", # noqa: E501
right_regex, # ,".*?"
r'(?:^|\n)(?P<quote>["\']).*?(?P=quote)(?:$|\n)'): # ".*?" (no delim, no space) # noqa: E501
regexp = re.compile(restr, re.DOTALL | re.MULTILINE)
matches = regexp.findall(sniff_line)
if matches:
break
# If it does not match the expression that was bugged,
# then this bug does not apply
if restr != right_regex:
return
groupindex = regexp.groupindex
# There is only one end of the string
assert len(matches) == 1
m = matches[0]
n = groupindex['quote'] - 1
quote = m[n]
n = groupindex['delim'] - 1
delim = m[n]
n = groupindex['space'] - 1
space = bool(m[n])
dq_regexp = re.compile(
rf"(({re.escape(delim)})|^)\W*{quote}[^{re.escape(delim)}\n]*{quote}[^{re.escape(delim)}\n]*{quote}\W*(({re.escape(delim)})|$)", re.MULTILINE # noqa: E501
)
doublequote = bool(dq_regexp.search(sniff_line))
dialect.quotechar = quote
if delim in delimiters:
dialect.delimiter = delim
dialect.doublequote = doublequote
dialect.skipinitialspace = space
def split_data_line(line, dialect=None):
delimiters = ",\t"
# This can not be done in a per reader basis, and relational fields
# can be HUGE
csv.field_size_limit(int(ctypes.c_ulong(-1).value // 2))
# Remove the line end if any
if line[-1] == '\n':
line = line[:-1]
# Remove potential trailing whitespace
line = line.strip()
sniff_line = line
# Add a delimiter if none is present, so that the csv.Sniffer
# does not complain for a single-field CSV.
if not any(d in line for d in delimiters):
sniff_line += ","
if dialect is None:
dialect = csv.Sniffer().sniff(sniff_line, delimiters=delimiters)
workaround_csv_sniffer_bug_last_field(sniff_line=sniff_line,
dialect=dialect,
delimiters=delimiters)
row = next(csv.reader([line], dialect))
return row, dialect
# --------------
# Parsing header
# --------------
def tokenize_attribute(iterable, attribute):
"""Parse a raw string in header (e.g., starts by @attribute).
Given a raw string attribute, try to get the name and type of the
attribute. Constraints:
* The first line must start with @attribute (case insensitive, and
space like characters before @attribute are allowed)
* Works also if the attribute is spread on multilines.
* Works if empty lines or comments are in between
Parameters
----------
attribute : str
the attribute string.
Returns
-------
name : str
name of the attribute
value : str
value of the attribute
next : str
next line to be parsed
Examples
--------
If attribute is a string defined in python as r"floupi real", will
return floupi as name, and real as value.
>>> from scipy.io.arff._arffread import tokenize_attribute
>>> iterable = iter([0] * 10) # dummy iterator
>>> tokenize_attribute(iterable, r"@attribute floupi real")
('floupi', 'real', 0)
If attribute is r"'floupi 2' real", will return 'floupi 2' as name,
and real as value.
>>> tokenize_attribute(iterable, r" @attribute 'floupi 2' real ")
('floupi 2', 'real', 0)
"""
sattr = attribute.strip()
mattr = r_attribute.match(sattr)
if mattr:
# atrv is everything after @attribute
atrv = mattr.group(1)
if r_comattrval.match(atrv):
name, type = tokenize_single_comma(atrv)
next_item = next(iterable)
elif r_wcomattrval.match(atrv):
name, type = tokenize_single_wcomma(atrv)
next_item = next(iterable)
else:
# Not sure we should support this, as it does not seem supported by
# weka.
raise ValueError("multi line not supported yet")
else:
raise ValueError(f"First line unparsable: {sattr}")
attribute = to_attribute(name, type)
if type.lower() == 'relational':
next_item = read_relational_attribute(iterable, attribute, next_item)
# raise ValueError("relational attributes not supported yet")
return attribute, next_item
def tokenize_single_comma(val):
# XXX we match twice the same string (here and at the caller level). It is
# stupid, but it is easier for now...
m = r_comattrval.match(val)
if m:
try:
name = m.group(1).strip()
type = m.group(2).strip()
except IndexError as e:
raise ValueError("Error while tokenizing attribute") from e
else:
raise ValueError(f"Error while tokenizing single {val}")
return name, type
def tokenize_single_wcomma(val):
# XXX we match twice the same string (here and at the caller level). It is
# stupid, but it is easier for now...
m = r_wcomattrval.match(val)
if m:
try:
name = m.group(1).strip()
type = m.group(2).strip()
except IndexError as e:
raise ValueError("Error while tokenizing attribute") from e
else:
raise ValueError(f"Error while tokenizing single {val}")
return name, type
def read_relational_attribute(ofile, relational_attribute, i):
"""Read the nested attributes of a relational attribute"""
r_end_relational = re.compile(r'^@[Ee][Nn][Dd]\s*' +
relational_attribute.name + r'\s*$')
while not r_end_relational.match(i):
m = r_headerline.match(i)
if m:
isattr = r_attribute.match(i)
if isattr:
attr, i = tokenize_attribute(ofile, i)
relational_attribute.attributes.append(attr)
else:
raise ValueError(f"Error parsing line {i}")
else:
i = next(ofile)
i = next(ofile)
return i
def read_header(ofile):
"""Read the header of the iterable ofile."""
i = next(ofile)
# Pass first comments
while r_comment.match(i):
i = next(ofile)
# Header is everything up to DATA attribute ?
relation = None
attributes = []
while not r_datameta.match(i):
m = r_headerline.match(i)
if m:
isattr = r_attribute.match(i)
if isattr:
attr, i = tokenize_attribute(ofile, i)
attributes.append(attr)
else:
isrel = r_relation.match(i)
if isrel:
relation = isrel.group(1)
else:
raise ValueError(f"Error parsing line {i}")
i = next(ofile)
else:
i = next(ofile)
return relation, attributes
class MetaData:
"""Small container to keep useful information on a ARFF dataset.
Knows about attributes names and types.
Examples
--------
::
data, meta = loadarff('iris.arff')
# This will print the attributes names of the iris.arff dataset
for i in meta:
print(i)
# This works too
meta.names()
# Getting attribute type
types = meta.types()
Methods
-------
names
types
Notes
-----
Also maintains the list of attributes in order, i.e., doing for i in
meta, where meta is an instance of MetaData, will return the
different attribute names in the order they were defined.
"""
def __init__(self, rel, attr):
self.name = rel
self._attributes = {a.name: a for a in attr}
def __repr__(self):
msg = ""
msg += f"Dataset: {self.name}\n"
for i in self._attributes:
msg += f"\t{i}'s type is {self._attributes[i].type_name}"
if self._attributes[i].range:
msg += f", range is {str(self._attributes[i].range)}"
msg += '\n'
return msg
def __iter__(self):
return iter(self._attributes)
def __getitem__(self, key):
attr = self._attributes[key]
return (attr.type_name, attr.range)
def names(self):
"""Return the list of attribute names.
Returns
-------
attrnames : list of str
The attribute names.
"""
return list(self._attributes)
def types(self):
"""Return the list of attribute types.
Returns
-------
attr_types : list of str
The attribute types.
"""
attr_types = [self._attributes[name].type_name
for name in self._attributes]
return attr_types
def loadarff(f):
"""
Read an arff file.
The data is returned as a record array, which can be accessed much like
a dictionary of NumPy arrays. For example, if one of the attributes is
called 'pressure', then its first 10 data points can be accessed from the
``data`` record array like so: ``data['pressure'][0:10]``
Parameters
----------
f : file-like or str
File-like object to read from, or filename to open.
Returns
-------
data : record array
The data of the arff file, accessible by attribute names.
meta : `MetaData`
Contains information about the arff file such as name and
type of attributes, the relation (name of the dataset), etc.
Raises
------
ParseArffError
This is raised if the given file is not ARFF-formatted.
NotImplementedError
The ARFF file has an attribute which is not supported yet.
Notes
-----
This function should be able to read most arff files. Not
implemented functionality include:
* date type attributes
* string type attributes
It can read files with numeric and nominal attributes. It cannot read
files with sparse data ({} in the file). However, this function can
read files with missing data (? in the file), representing the data
points as NaNs.
Examples
--------
>>> from scipy.io import arff
>>> from io import StringIO
>>> content = \"\"\"
... @relation foo
... @attribute width numeric
... @attribute height numeric
... @attribute color {red,green,blue,yellow,black}
... @data
... 5.0,3.25,blue
... 4.5,3.75,green
... 3.0,4.00,red
... \"\"\"
>>> f = StringIO(content)
>>> data, meta = arff.loadarff(f)
>>> data
array([(5.0, 3.25, 'blue'), (4.5, 3.75, 'green'), (3.0, 4.0, 'red')],
dtype=[('width', '<f8'), ('height', '<f8'), ('color', '|S6')])
>>> meta
Dataset: foo
\twidth's type is numeric
\theight's type is numeric
\tcolor's type is nominal, range is ('red', 'green', 'blue', 'yellow', 'black')
"""
if hasattr(f, 'read'):
ofile = f
else:
ofile = open(f)
try:
return _loadarff(ofile)
finally:
if ofile is not f: # only close what we opened
ofile.close()
def _loadarff(ofile):
# Parse the header file
try:
rel, attr = read_header(ofile)
except ValueError as e:
msg = "Error while parsing header, error was: " + str(e)
raise ParseArffError(msg) from e
# Check whether we have a string attribute (not supported yet)
hasstr = False
for a in attr:
if isinstance(a, StringAttribute):
hasstr = True
meta = MetaData(rel, attr)
# XXX The following code is not great
# Build the type descriptor descr and the list of converters to convert
# each attribute to the suitable type (which should match the one in
# descr).
# This can be used once we want to support integer as integer values and
# not as numeric anymore (using masked arrays ?).
if hasstr:
# How to support string efficiently ? Ideally, we should know the max
# size of the string before allocating the numpy array.
raise NotImplementedError("String attributes not supported yet, sorry")
ni = len(attr)
def generator(row_iter, delim=','):
# TODO: this is where we are spending time (~80%). I think things
# could be made more efficiently:
# - We could for example "compile" the function, because some values
# do not change here.
# - The function to convert a line to dtyped values could also be
# generated on the fly from a string and be executed instead of
# looping.
# - The regex are overkill: for comments, checking that a line starts
# by % should be enough and faster, and for empty lines, same thing
# --> this does not seem to change anything.
# 'compiling' the range since it does not change
# Note, I have already tried zipping the converters and
# row elements and got slightly worse performance.
elems = list(range(ni))
dialect = None
for raw in row_iter:
# We do not abstract skipping comments and empty lines for
# performance reasons.
if r_comment.match(raw) or r_empty.match(raw):
continue
row, dialect = split_data_line(raw, dialect)
yield tuple([attr[i].parse_data(row[i]) for i in elems])
a = list(generator(ofile))
# No error should happen here: it is a bug otherwise
data = np.array(a, [(a.name, a.dtype) for a in attr])
return data, meta
|