niimpy.reading.read module¶
Read data from various formats, user entery point.
This module contains various functions read_* which load data from different formats into pandas.DataFrame:s. As a side effect, it provides the authoritative information on how incoming data is converted to dataframes.
- niimpy.reading.read.read_csv(filename, read_csv_options={}, add_group=None, tz=None)[source]¶
Read DataFrame from csv file
This will read data from a csv file and then process the result with niimpy.util.df_normalize.
- Parameters
- filenamestr
filename of csv file
- read_csv_options: dict
Dictionary of options to pandas.read_csv, if this is necessary for custom csv files.
- add_groupobject
If given, add a ‘group’ column with all values set to this.
- niimpy.reading.read.read_csv_string(string, tz=None)[source]¶
Parse a string containing CSV and return dataframe
This should not be used for serious reading of CSV from disk, but can be useful for tests and examples. Various CSV reading options are turned on in order to be better for examples:
Allow comments in the CSV file
Remove the datetime column (redundant with index but some older functions break without it, so default readers need to leave it).
- Parameters
- stringstring containing CSV file
- Returns
- df: pandas.DataFrame
- niimpy.reading.read.read_sqlite(filename, table, add_group=None, user=<class 'niimpy.reading.database.ALL'>, limit=None, offset=None, start=None, end=None, tz=None)[source]¶
Read DataFrame from sqlite3 database
This will read data from a sqlite3 file, taking sensor data in a given table, and optionally apply various limits.
- Parameters
- filenamestr
filename of sqlite3 database
- tablestr
table name of data within the database
- add_groupobject
If given, add a ‘group’ column with all values set to this.
- userstr or database.ALL, optional
If given, return only data matching this user (based an column ‘user’)
- limitint, optional
If given, return only this many rows
- offsetint, optional
When used with limit, skip this many lines at the beginning
- startint or float or str or datetime.datetime, optional
If given, limit to this starting time. Formats can be int/float (unixtime), string (parsed with dateutil.parser.parser, or datetime.datetime.
- endint or float or str or datetime.datetime, optional
Same meaning as ‘start’, but for end time