Coverage for ebcpy/data_types.py: 96%
250 statements
« prev ^ index » next coverage.py v7.4.4, created at 2025-08-26 09:12 +0000
« prev ^ index » next coverage.py v7.4.4, created at 2025-08-26 09:12 +0000
1"""
2This module provides useful classes for all ebcpy.
3Every data_type class should include every parameter
4other classes like optimization etc. may need. The checking
5of correct input is especially relevant here as the correct
6format of data-types will prevent errors during simulations,
7optimization etc.
8"""
10import os
11import warnings
12from pathlib import Path
13from typing import List, Union, Any, TYPE_CHECKING
14from datetime import datetime
15from pandas.core.internals import BlockManager
16import pandas as pd
17import numpy as np
18import ebcpy.modelica.simres as sr
20from ebcpy.utils import get_names
21from ebcpy import preprocessing
23# pylint: disable=I1101
24# pylint: disable=too-many-ancestors
26__all__ = ['TimeSeries',
27 'TimeSeriesData',
28 'numeric_index_dtypes',
29 'load_time_series_data',
30 'index_is_numeric',
31 'datetime_indexes']
33numeric_index_dtypes = [
34 pd.Index([], dtype=dtype).dtype for dtype in
35 ["int8", "int16", "int32", "int64",
36 "uint8", "uint16", "uint32", "uint64",
37 "float32", "float64"]
38]
40datetime_indexes = [
41 pd.DatetimeIndex
42]
45def index_is_numeric(index: pd.Index):
46 """Check if pandas Index is numeric"""
47 return isinstance(index, pd.RangeIndex) or index.dtype in numeric_index_dtypes
50@pd.api.extensions.register_dataframe_accessor("tsd")
51class TimeSeriesAccessor:
52 """
53 Pandas DataFrame accessor for time series functionality.
54 Access using df.tsd.*
55 """
57 def __init__(self, pandas_obj):
58 self._obj = pandas_obj
59 self._filepath = None
61 @property
62 def filepath(self):
63 """Get the filepath associated with the time series data"""
64 return self._filepath
66 @filepath.setter
67 def filepath(self, filepath):
68 """Set the filepath associated with the time series data"""
69 self._filepath = Path(filepath) if filepath else None
71 def save(self, filepath: str = None, **kwargs) -> None:
72 """
73 Save the current time-series-data into the given file-format.
74 Currently supported are .hdf, which is an easy and fast storage,
75 and, .csv is supported as an easy-readable option.
76 Also, .parquet, and with additional compression .parquet.COMPRESSION_NAME
77 are supported.
79 :param str,os.path.normpath filepath:
80 Filepath were to store the data. Either .hdf, .csv, .parquet
81 or .parquet.COMPRESSION_NAME has to be the file-ending.
82 Default is current filepath of class.
83 :keyword str key:
84 Necessary keyword-argument for saving a .hdf-file.
85 Specifies the key of the table in the .hdf-file.
86 :keyword str sep:
87 Separator used for saving as .csv. Default is ','.
88 :keyword str engine:
89 Chose the engine for reading .parquet files. Default is 'pyarrow'
90 Other option is 'fastparquet' (python>=3.9).
91 """
92 # Set filepath if not given
93 if filepath is None:
94 if self.filepath is None:
95 raise FileNotFoundError(
96 "TimeSeriesData has neither a filepath stored in tsd "
97 "accessor nor did you provide a filepath were to store the data."
98 )
99 filepath = self.filepath
100 else:
101 filepath = Path(filepath)
103 # Check if filepath is still None
104 if filepath is None:
105 raise ValueError("No filepath specified and no default filepath is set.")
107 # Save based on file suffix
108 if filepath.suffix == ".hdf":
109 if "key" not in kwargs:
110 raise KeyError("Argument 'key' must be specified to save a .hdf file")
111 self._obj.to_hdf(filepath, key=kwargs.get("key"))
112 elif filepath.suffix == ".csv":
113 self._obj.to_csv(filepath, sep=kwargs.get("sep", ","))
114 elif ".parquet" in filepath.name:
115 parquet_split = filepath.name.split(".parquet")
116 self._obj.to_parquet(
117 filepath, engine=kwargs.get('engine', 'pyarrow'),
118 compression=parquet_split[-1][1:] if parquet_split[-1] else None,
119 index=True
120 )
121 else:
122 raise TypeError("Given file-format is not supported."
123 "You can only store time series data as .hdf, .csv, .parquet, "
124 "and .parquet.COMPRESSION_NAME with additional compression options")
126 def to_datetime_index(self, unit_of_index="s", origin=datetime.now(), inplace=True):
127 """
128 Convert the current index to a datetime index using
129 ebcpy.preprocessing.convert_index_to_datetime_index()
131 :param str unit_of_index: default 's'
132 The unit of the given index. Used to convert to
133 total_seconds later on.
134 :param datetime.datetime origin:
135 The reference datetime object for the first index.
136 Default is the current system time.
137 :param bool inplace:
138 If True, performs operation inplace and returns None.
139 :return: df
140 Copy of DataFrame with correct index for usage in this
141 framework.
142 """
143 return preprocessing.convert_index_to_datetime_index(
144 df=self._obj,
145 unit_of_index=unit_of_index,
146 origin=origin,
147 inplace=inplace
148 )
150 def to_float_index(self, offset=0, inplace=True):
151 """
152 Convert the current index to a float based index using
153 ebcpy.preprocessing.convert_datetime_index_to_float_index()
155 :param float offset:
156 Offset in seconds
157 :param bool inplace:
158 If True, performs operation inplace and returns None.
159 :return: pd.DataFrame df:
160 DataFrame with correct index.
161 """
162 if not isinstance(self._obj.index, pd.DatetimeIndex):
163 if inplace:
164 return None
165 return self._obj
167 return preprocessing.convert_datetime_index_to_float_index(
168 df=self._obj,
169 offset=offset,
170 inplace=inplace
171 )
173 def clean_and_space_equally(self, desired_freq, inplace=True):
174 """
175 Call to the preprocessing function
176 ebcpy.preprocessing.clean_and_space_equally_time_series()
177 See the docstring of this function to know what is happening.
179 :param str desired_freq:
180 Frequency to determine number of elements in processed dataframe.
181 Options are for example:
182 - s: second-based
183 - 5s: Every 5 seconds
184 - 6min: Every 6 minutes
185 This also works for h, d, m, y, ms etc.
186 :param bool inplace:
187 If True, performs operation inplace and returns None.
188 :return: pd.DataFrame
189 Cleaned and equally spaced data-frame
190 """
191 df = preprocessing.clean_and_space_equally_time_series(
192 df=self._obj,
193 desired_freq=desired_freq
194 )
195 if inplace:
196 self._obj = df
197 return None
198 return df
200 def low_pass_filter(self, crit_freq, filter_order, variable):
201 """
202 Call to the preprocessing function
203 ebcpy.preprocessing.low_pass_filter()
204 See the docstring of this function to know what is happening.
206 :param float crit_freq:
207 The critical frequency or frequencies.
208 :param int filter_order:
209 The order of the filter
210 :param str variable:
211 The variable name to apply the filter to
212 :return: numpy.ndarray
213 Filtered data
214 """
215 return preprocessing.low_pass_filter(
216 data=self._obj[variable].to_numpy(),
217 filter_order=filter_order,
218 crit_freq=crit_freq
219 )
221 def moving_average(self, window, variable):
222 """
223 Call to the preprocessing function
224 ebcpy.preprocessing.moving_average()
225 See the docstring of this function to know what is happening.
227 :param int window:
228 sample rate of input
229 :param str variable:
230 The variable name to apply the filter to
231 :return: numpy.ndarray
232 Moving average result
233 """
234 return preprocessing.moving_average(
235 data=self._obj[variable].to_numpy(),
236 window=window,
237 )
239 def get_variable_names(self, patterns: Union[str, List[str]] = None) -> List[str]:
240 """
241 Return an alphabetically sorted list of variable names, optionally filtered by patterns.
243 By default, returns all column names found in the DataFrame, sorted alphabetically.
244 If `patterns` is provided, only names matching one or more of the given
245 literal strings or glob-style patterns (where `*` matches any sequence of characters)
246 will be returned.
248 :param patterns:
249 - A single string or list of strings.
250 - Each entry may be an exact variable name, or a pattern containing `*` as a wildcard.
251 - If None, all variable names are returned.
252 :return:
253 A list of matching variable names, in alphabetical order.
254 :raises KeyError:
255 If any literal name or pattern does not match at least one variable in the DataFrame.
257 Example:
258 # return all wall temperatures at any layer
259 df.tsd.get_variable_names("*wall.layer[*].T")
260 ["wall.layer[1].T", "wall.layer[2].T", "wall.layer[3].T"]
261 """
262 all_names = sorted(self._obj.columns.get_level_values(0).unique())
263 if patterns is None:
264 return all_names
265 return get_names(all_names, patterns)
267 def number_lines_totally_na(self):
268 """
269 Returns the number of rows in the given dataframe
270 that are filled with NaN-values.
271 """
272 return preprocessing.number_lines_totally_na(self._obj)
274 @property
275 def frequency(self):
276 """
277 The frequency of the time series data.
278 Returns's the mean and the standard deviation of
279 the index.
281 :returns:
282 float: Mean value
283 float: Standard deviation
284 """
285 return preprocessing.get_df_index_frequency_mean_and_std(
286 df_index=self._obj.index
287 )
290class TimeSeriesData(pd.DataFrame):
291 """
292 Most data related to energy and building
293 climate related problems is time-variant.
295 Class for handling time series data using a pandas dataframe.
296 This class works file-based and makes the import of different
297 file-types into a pandas DataFrame more user-friendly.
298 Furthermore, functions to support multi-indexing are provided to
299 efficiently handle variable passed processing and provide easy
300 visualization and preprocessing access.
302 :param str,os.path.normpath,pd.DataFrame data:
303 Filepath ending with either .hdf, .mat, .csv, .parquet,
304 or .parquet.COMPRESSION_NAME containing
305 time-dependent data to be loaded as a pandas.DataFrame.
306 Alternative option is to pass a DataFrame directly.
307 :keyword str key:
308 Name of the table in a .hdf-file if the file
309 contains multiple tables.
310 :keyword str sep:
311 separator for the use of a csv file. If none is provided,
312 a comma (",") is used as a default value.
313 See pandas.read_csv() docs for further information.
314 :keyword int, list header:
315 Header columns for .csv files.
316 See pandas.read_csv() docs for further information.
317 Default is first row (0).
318 :keyword int,str index_col:
319 Column to be used as index in .csv files.
320 See pandas.read_csv() docs for further information.
321 Default is first column (0).
322 :keyword str sheet_name:
323 Name of the sheet you want to load data from. Required keyword
324 argument when loading a xlsx-file.
325 :keyword str default_tag:
326 Which value to use as tag. Default is 'raw'
327 :keyword str engine:
328 Chose the engine for reading .parquet files. Default is 'pyarrow'
329 Other option is 'fastparquet' (python>=3.9).
330 :keyword list variable_names:
331 List of variable names to load from .mat file. If you
332 know which variables you want to plot, this may speed up
333 loading significantly, and reduce memory size drastically.
334 You can also supply wildcard patterns (e.g. "*wall.layer[*].T", etc.)
335 to match multiple variables at once.
337 Examples:
339 First let's see the usage for a common dataframe.
341 >>> import numpy as np
342 >>> from ebcpy import TimeSeriesData
343 >>> tsd = TimeSeriesData({"my_variable": np.random.rand(5)})
344 >>> tsd.to_datetime_index()
345 >>> tsd.save("my_new_data.csv")
347 Now, let's load the recently created file.
349 >>> tsd = TimeSeriesData("my_new_data.csv")
350 """
352 # normal properties
353 _metadata = [
354 "_filepath",
355 "_loader_kwargs",
356 "_default_tag",
357 "_multi_col_names"
358 ]
360 def __init__(self, data: Union[str, Any], use_multicolumn: bool = False, **kwargs):
361 """Initialize class-objects and check correct input."""
362 warnings.warn(
363 "TimeSeriesData will be deprecated in the next major release. "
364 "Instead, use 'load_time_series_data' to load files etc. as pd.DataFrame "
365 "and use the 'tsd' accessor to access useful time-series-related functions "
366 "as before with TimeSeriesData.", FutureWarning
367 )
368 if use_multicolumn:
369 warnings.warn(
370 "All multicolumn support will be removed in the next major release", FutureWarning
371 )
372 # Initialize as default
373 self._filepath = None
374 self._loader_kwargs = {}
375 self._multi_col_names = ["Variables", "Tags"]
376 self._default_tag = kwargs.pop("default_tag", "raw")
377 if not isinstance(self._default_tag, str):
378 raise TypeError(f"Invalid type for default_tag! Expected 'str' but "
379 f"received {type(self._default_tag)}")
381 # Two possibles inputs. first argument is actually data provided by pandas
382 # and kwargs hold further information or is it an actual filepath.
383 if isinstance(data, BlockManager):
384 super().__init__(data=data)
385 return
387 if not isinstance(data, (str, Path)):
388 _df_loaded = pd.DataFrame(data=data,
389 index=kwargs.get("index", None),
390 columns=kwargs.get("columns", None),
391 dtype=kwargs.get("dtype", None),
392 copy=kwargs.get("copy", False))
393 else:
394 file = Path(data)
395 self._loader_kwargs = kwargs.copy()
396 _df_loaded = _load_df_from_file(file=file, **self._loader_kwargs)
397 _df_loaded.tsd.filepath = file
398 self._filepath = file
400 if _df_loaded.columns.nlevels == 1:
401 # Check if first level is named Tags.
402 # If so, don't create MultiIndex-DF as the method is called by the pd constructor
403 if _df_loaded.columns.name != self._multi_col_names[1] and use_multicolumn:
404 multi_col = pd.MultiIndex.from_product(
405 [_df_loaded.columns, [self._default_tag]],
406 names=self._multi_col_names
407 )
408 _df_loaded.columns = multi_col
410 elif _df_loaded.columns.nlevels == 2:
411 if _df_loaded.columns.names != self._multi_col_names and use_multicolumn:
412 raise TypeError("Loaded dataframe has a different 2-Level "
413 "header format than it is supported by this "
414 "class. The names have to match.")
415 else:
416 raise TypeError("Only DataFrames with Multi-Columns with 2 "
417 "Levels are supported by this class.")
419 super().__init__(_df_loaded)
421 @property
422 def _constructor(self):
423 """Overwrite constructor method according to:
424 https://pandas.pydata.org/pandas-docs/stable/development/extending.html#extending-subclassing-pandas"""
425 return TimeSeriesData
427 @property
428 def _constructor_sliced(self):
429 """Overwrite constructor method according to:
430 https://pandas.pydata.org/pandas-docs/stable/development/extending.html#extending-subclassing-pandas"""
431 return TimeSeries
433 @property
434 def filepath(self) -> str:
435 """Get the filepath associated with the time series data"""
436 return self._filepath
438 @filepath.setter
439 def filepath(self, filepath: str):
440 """Set the filepath associated with the time series data"""
441 self._filepath = Path(filepath)
442 self.tsd.filepath = self._filepath
444 @property
445 def default_tag(self) -> str:
446 """Get the default of time series data object"""
447 return self._default_tag
449 @default_tag.setter
450 def default_tag(self, tag: str) -> None:
451 """Set the default_tag of the time series data object
452 :param tag: new tag
453 :type tag: String
454 """
455 if not isinstance(tag, str):
456 raise TypeError(f"Invalid type for default_tag! Expected 'str' but "
457 f"received {type(tag)}")
458 if tag not in self.get_tags():
459 raise KeyError(f"Tag '{tag}' does not exist for current data set!"
460 f"\n Available tags: {self.get_tags()}")
461 self._default_tag = tag
463 def save(self, filepath: str = None, **kwargs) -> None:
464 """
465 Save the current time-series-data into the given file-format.
466 Currently supported are .hdf, which is an easy and fast storage,
467 and, .csv is supported as an easy-readable option.
468 Also, .parquet, and with additional compression .parquet.COMPRESSION_NAME
469 are supported. Compressions could be gzip, brotli or snappy. For all possible
470 compressions see the documentation of the parquet engines.
471 For a small comparison of these data formats see https://github.com/RWTH-EBC/ebcpy/issues/81
473 :param str,os.path.normpath filepath:
474 Filepath were to store the data. Either .hdf, .csv, .parquet
475 or .parquet.COMPRESSION_NAME has to be the file-ending.
476 Default is current filepath of class.
477 :keyword str key:
478 Necessary keyword-argument for saving a .hdf-file.
479 Specifies the key of the table in the .hdf-file.
480 :keyword str sep:
481 Separator used for saving as .csv. Default is ','.
482 :keyword str engine:
483 Chose the engine for reading .parquet files. Default is 'pyarrow'
484 Other option is 'fastparquet' (python>=3.9).
485 :return:
486 """
487 # If new settings are needed, update existing ones
488 self._loader_kwargs.update(kwargs)
489 self.tsd.save(filepath, **kwargs)
491 def to_df(self, force_single_index=False):
492 """
493 Return the dataframe version of the current TimeSeriesData object.
494 If all tags are equal, the tags are dropped.
495 Else, the object is just converted.
497 :param bool force_single_index:
498 If True (not the default), the conversion to a standard
499 DataFrame with a single index column (only variable names)
500 is only done if no variable contains multiple tags.
501 """
502 if len(self.get_variables_with_multiple_tags()) == 0:
503 if self._is_old_multicolumn_format:
504 return pd.DataFrame(self.droplevel(1, axis=1))
505 return pd.DataFrame(self)
506 if force_single_index:
507 raise IndexError(
508 "Can't automatically drop all tags "
509 "as the following variables contain multiple tags: "
510 f"{' ,'.join(self.get_variables_with_multiple_tags())}. "
511 )
512 return pd.DataFrame(self)
514 def get_variable_names(self, patterns: Union[str, List[str]] = None) -> List[str]:
515 """
516 Return an alphabetically sorted list of variable names, optionally filtered by patterns.
518 By default, returns all variable names found in the first level of the DataFrame's
519 column MultiIndex, sorted alphabetically. If `patterns` is provided, only names
520 matching one or more of the given literal strings or glob-style patterns
521 (where `*` matches any sequence of characters) will be returned.
523 :param patterns:
524 - A single string or list of strings.
525 - Each entry may be an exact variable name, or a pattern containing `*` as a wildcard.
526 - If None, all variable names are returned.
527 :return:
528 A list of matching variable names, in alphabetical order.
529 :raises KeyError:
530 If any literal name or pattern does not match at least one variable in the DataFrame.
532 Example:
533 # return all wall temperatures at any layer
534 tsd.get_variable_names("*wall.layer[*].T")
535 ["wall.layer[1].T", "wall.layer[2].T", "wall.layer[3].T"]
536 """
537 return self.tsd.get_variable_names(patterns)
539 def get_variables_with_multiple_tags(self) -> List[str]:
540 """
541 Return an alphabetically sorted list of all variables
542 that contain more than one tag.
544 :return: List[str]
545 """
546 var_names = self.columns.get_level_values(0)
547 return sorted(var_names[var_names.duplicated()])
549 def get_tags(self, variable: str = None) -> List[str]:
550 """
551 Return an alphabetically sorted list of all tags
553 :param str variable:
554 If given, tags of this variable are returned
556 :return: List[str]
557 """
558 if not self._is_old_multicolumn_format:
559 raise KeyError("You can't get tags for a TimeSeriesData object created with use_multicolumn=False!")
560 if variable:
561 tags = self.loc[:, variable].columns
562 return sorted(tags)
563 return sorted(self.columns.get_level_values(1).unique())
565 @property
566 def _is_old_multicolumn_format(self):
567 """
568 Helper function to check if the old multicolumn format is used.
569 """
570 return isinstance(self.columns, pd.MultiIndex)
572 def get_columns_by_tag(self,
573 tag: str,
574 variables: list = None,
575 return_type: str = 'pandas',
576 drop_level: bool = False):
577 """
578 Returning all columns with defined tag in the form of ndarray.
580 :param str tag:
581 Define the tag which return columns have to
582 match.
583 :param list variables:
584 Besides the given tag, specify the
585 variables names matching the return criteria as well.
586 :param boolean drop_level:
587 If tag should be included in the response.
588 Default is True.
589 :param str return_type:
590 Return format. Options are:
591 - pandas (pd.series)
592 - numpy, scipy, sp, and np (np.array)
593 - control (transposed np.array)
594 :return: ndarray of input signals
595 """
596 if not self._is_old_multicolumn_format:
597 raise KeyError("You can't get tags for a TimeSeriesData object created with use_multicolumn=False!")
599 # Extract columns
600 if variables:
601 _ret = self.loc[:, variables]
602 else:
603 _ret = self
605 _ret = _ret.xs(tag, axis=1, level=1, drop_level=drop_level)
607 # Return based on the given return_type
608 if return_type.lower() == 'pandas':
609 return _ret
610 if return_type.lower() in ['numpy', 'scipy', 'sp', 'np']:
611 return _ret.to_numpy()
612 if return_type.lower() == 'control':
613 return _ret.to_numpy().transpose()
614 raise TypeError("Unknown return type")
616 def to_datetime_index(self, unit_of_index="s", origin=datetime.now(), inplace: bool = True):
617 """
618 Convert the current index to a float based index using
619 ebcpy.preprocessing.convert_index_to_datetime_index()
621 :param str unit_of_index: default 's'
622 The unit of the given index. Used to convert to
623 total_seconds later on.
624 :param datetime.datetime origin:
625 The reference datetime object for the first index.
626 Default is the current system time.
627 :param bool inplace:
628 If True, performs operation inplace and returns None.
629 :return: df
630 Copy of DataFrame with correct index for usage in this
631 framework.
633 """
634 return self.tsd.to_datetime_index(unit_of_index, origin, inplace)
636 def to_float_index(self, offset=0, inplace: bool = True):
637 """
638 Convert the current index to a float based index using
639 ebcpy.preprocessing.convert_datetime_index_to_float_index()
641 :param float offset:
642 Offset in seconds
643 :param bool inplace:
644 If True, performs operation inplace and returns None.
645 :return: pd.DataFrame df:
646 DataFrame with correct index.
647 """
648 return self.tsd.to_float_index(offset, inplace)
650 def clean_and_space_equally(self, desired_freq, inplace: bool = True):
651 """
652 Call to the preprocessing function
653 ebcpy.preprocessing.clean_and_space_equally_time_series()
654 See the docstring of this function to know what is happening.
656 :param str desired_freq:
657 Frequency to determine number of elements in processed dataframe.
658 Options are for example:
659 - s: second-based
660 - 5s: Every 5 seconds
661 - 6min: Every 6 minutes
662 This also works for h, d, m, y, ms etc.
663 :param bool inplace:
664 If True, performs operation inplace and returns None.
665 :return: pd.DataFrame
666 Cleaned and equally spaced data-frame
667 """
668 return self.tsd.clean_and_space_equally(desired_freq, inplace)
670 def low_pass_filter(self, crit_freq, filter_order, variable,
671 tag=None, new_tag="low_pass_filter"):
672 """
673 Call to the preprocessing function
674 ebcpy.preprocessing.low_pass_filter()
675 See the docstring of this function to know what is happening.
676 If the old multicolumn format is used, the result is stored in the
677 multicolumn header with the `new_tag`.
679 :param float crit_freq:
680 The critical frequency or frequencies.
681 :param int filter_order:
682 The order of the filter
683 :param str variable:
684 The variable name to apply the filter to
685 :param str tag:
686 If this variable has more than one tag, specify which one
687 :param str new_tag:
688 The new tag to pass to the variable.
689 Default is 'low_pass_filter'
690 """
691 result = self.tsd.low_pass_filter(crit_freq, filter_order, self._possibly_get_variable_and_tag(variable, tag))
692 if self._is_old_multicolumn_format:
693 self.loc[:, (variable, new_tag)] = result
694 else:
695 return result
697 def moving_average(self, window, variable,
698 tag=None, new_tag="moving_average"):
699 """
700 Call to the preprocessing function
701 ebcpy.preprocessing.moving_average()
702 See the docstring of this function to know what is happening.
703 If the old multicolumn format is used, the result is stored in the
704 multicolumn header with the `new_tag`.
706 :param int window:
707 sample rate of input
708 :param str variable:
709 The variable name to apply the filter to
710 :param str tag:
711 If this variable has more than one tag, specify which one
712 :param str new_tag:
713 The new tag to pass to the variable.
714 Default is 'low_pass_filter'
715 """
716 result = self.tsd.moving_average(window, self._possibly_get_variable_and_tag(variable, tag))
717 if self._is_old_multicolumn_format:
718 self.loc[:, (variable, new_tag)] = result
719 else:
720 return result
722 def _possibly_get_variable_and_tag(self, variable: str, tag: str = None):
723 """
724 Helper function to get numpy array based on variable and possible tag name,
725 depending on whether multicolumn is used or not.
727 :param str variable:
728 The variable name to apply the filter to
729 :param str tag:
730 If this variable has more than one tag, specify which one
732 """
733 if tag is None:
734 return variable
735 if self._is_old_multicolumn_format:
736 return (variable, tag)
737 return variable
739 def number_lines_totally_na(self):
740 """
741 Returns the number of rows in the given dataframe
742 that are filled with NaN-values.
743 """
744 return self.tsd.number_lines_totally_na()
746 @property
747 def frequency(self):
748 """
749 The frequency of the time series data.
750 Returns's the mean and the standard deviation of
751 the index.
753 :returns:
754 float: Mean value
755 float: Standard deviation
756 """
757 return self.tsd.frequency
760class TimeSeries(pd.Series):
761 """Overwrites pd.Series to enable correct slicing
762 and expansion in the TimeSeriesData class
764 .. versionadded:: 0.1.7
765 """
767 @property
768 def _constructor(self):
769 """Overwrite constructor method according to:
770 https://pandas.pydata.org/pandas-docs/stable/development/extending.html#extending-subclassing-pandas"""
771 return TimeSeries
773 @property
774 def _constructor_expanddim(self):
775 """Overwrite constructor method according to:
776 https://pandas.pydata.org/pandas-docs/stable/development/extending.html#extending-subclassing-pandas"""
777 return TimeSeriesData
780def get_keys_of_hdf_file(filepath):
781 """
782 Find all keys in a given hdf-file.
784 :param str,os.path.normpath filepath:
785 Path to the .hdf-file
786 :return: list
787 List with all keys in the given file.
788 """
789 # pylint: disable=import-outside-toplevel
790 try:
791 import h5py
792 with h5py.File(filepath, 'r') as hdf_file:
793 return list(hdf_file.keys())
794 except ImportError:
795 return ["ERROR: Could not obtain keys as h5py is not installed"]
798def load_time_series_data(data: Union[str, Any], **kwargs) -> pd.DataFrame:
799 """
800 Load time series data from various sources into a pandas DataFrame with
801 custom time series accessor methods available via .tsd property.
803 :param str,os.path.normpath,pd.DataFrame data:
804 Filepath ending with either .hdf, .mat, .csv, .parquet,
805 or .parquet.COMPRESSION_NAME containing
806 time-dependent data to be loaded as a pandas.DataFrame.
807 Alternative option is to pass a DataFrame directly.
808 :keyword str key:
809 Name of the table in a .hdf-file if the file
810 contains multiple tables.
811 :keyword str sep:
812 separator for the use of a csv file. If none is provided,
813 a comma (",") is used as a default value.
814 See pandas.read_csv() docs for further information.
815 :keyword int, list header:
816 Header columns for .csv files.
817 See pandas.read_csv() docs for further information.
818 Default is first row (0).
819 :keyword int,str index_col:
820 Column to be used as index in .csv files.
821 See pandas.read_csv() docs for further information.
822 Default is first column (0).
823 :keyword str sheet_name:
824 Name of the sheet you want to load data from. Required keyword
825 argument when loading a xlsx-file.
826 :keyword str engine:
827 Chose the engine for reading .parquet files. Default is 'pyarrow'
828 Other option is 'fastparquet' (python>=3.9).
829 :keyword list variable_names:
830 List of variable names to load from .mat file. If you
831 know which variables you want to plot, this may speed up
832 loading significantly, and reduce memory size drastically.
833 You can also supply wildcard patterns (e.g. "*wall.layer[*].T", etc.)
834 to match multiple variables at once.
835 :return: pd.DataFrame
836 DataFrame with custom .tsd accessor containing time series functionality
838 Examples:
840 Create a DataFrame with random data:
842 >>> import numpy as np
843 >>> from ebcpy import load_time_series_data
844 >>> df = load_time_series_data({"my_variable": np.random.rand(5)})
845 >>> df.tsd.to_datetime_index()
846 >>> df.tsd.save("my_new_data.csv")
848 Now, let's load the recently created file:
850 >>> df = load_time_series_data("my_new_data.csv")
851 """
852 if isinstance(data, pd.DataFrame):
853 df = data.copy()
854 elif not isinstance(data, (str, Path)):
855 df = pd.DataFrame(data=data,
856 index=kwargs.get("index", None),
857 columns=kwargs.get("columns", None),
858 dtype=kwargs.get("dtype", None),
859 copy=kwargs.get("copy", False))
860 else:
861 # Load from file
862 file = Path(data)
863 df = _load_df_from_file(file=file, **kwargs)
864 df.tsd.filepath = file
866 return df
869def _load_df_from_file(file, **kwargs):
870 """
871 Function to load a given filepath into a dataframe
873 :param Path file: File path to load
874 :param kwargs: Additional loading parameters
875 :return: pd.DataFrame
876 """ # Check whether the file exists
877 if not os.path.isfile(file):
878 raise FileNotFoundError(
879 f"The given filepath {file} could not be opened")
881 # Open based on file suffix.
882 # Currently, hdf, csv, and Modelica result files (mat) are supported.
883 if file.suffix == ".hdf":
884 # Load the current file as a hdf to a dataframe.
885 # As specifying the key can be a problem, the user will
886 # get all keys of the file if one is necessary but not provided.
887 key = kwargs.get("key")
888 if key == "":
889 key = None # Avoid cryptic error in pandas by converting empty string to None
890 try:
891 df = pd.read_hdf(file, key=key)
892 except (ValueError, KeyError) as error:
893 keys = ", ".join(get_keys_of_hdf_file(file))
894 raise KeyError(f"key must be provided when HDF5 file contains multiple datasets. "
895 f"Here are all keys in the given hdf-file: {keys}") from error
896 elif file.suffix == ".csv":
897 # Check if file was previously a TimeSeriesData object
898 with open(file, "r") as _f:
899 lines = [_f.readline() for _ in range(2)]
900 # Backwards compatible assumption: Users never changed '_multi_col_names'
901 if (lines[0].startswith("Variables") and
902 lines[1].startswith("Tags")):
903 _hea_def = [0, 1]
904 else:
905 _hea_def = 0
907 df = pd.read_csv(
908 file,
909 sep=kwargs.get("sep", ","),
910 index_col=kwargs.get("index_col", 0),
911 header=kwargs.get("header", _hea_def)
912 )
913 elif file.suffix == ".mat":
914 df = sr.mat_to_pandas(
915 fname=file,
916 with_unit=False,
917 names=kwargs.get("variable_names")
918 )
919 elif file.suffix in ['.xlsx', '.xls', '.odf', '.ods', '.odt']:
920 sheet_name = kwargs.get("sheet_name")
921 if sheet_name is None:
922 raise KeyError("sheet_name is a required keyword argument to load xlsx-files."
923 "Please pass a string to specify the name "
924 "of the sheet you want to load.")
925 df = pd.read_excel(io=file, sheet_name=sheet_name)
926 elif ".parquet" in file.name:
927 df = pd.read_parquet(path=file, engine=kwargs.get('engine', 'pyarrow'))
928 else:
929 raise TypeError("Only .hdf, .csv, .xlsx and .mat are supported!")
930 if not isinstance(df.index, tuple(datetime_indexes)) and not index_is_numeric(df.index):
931 try:
932 df.index = pd.DatetimeIndex(df.index)
933 except Exception as err:
934 raise IndexError(
935 f"Given data has index of type {type(df.index)}. "
936 f"Currently only numeric indexes and the following are supported:"
937 f"{' ,'.join([str(idx) for idx in [pd.RangeIndex] + datetime_indexes])} "
938 f"Automatic conversion to pd.DateTimeIndex failed"
939 f"see error above."
940 ) from err
941 return df