Input/Output

API	Description
`load_geolife_trajectories`	Load Microsoft GeoLife `.plt` trajectory files into a `TrajDataFrame`.
`read`	Read a trajectory table from disk and return a `TrajDataFrame`.
`write`	Write a trajectory dataframe to disk.

`fastmob.io.file.load_geolife_trajectories(path, user_ids=None, **kwargs)`

Load Microsoft GeoLife .plt trajectory files into a TrajDataFrame.

Parameters:

Name	Type	Description	Default
`path`	`str`	local path of the directory 'Geolife Trajectories 1.3/'	required
`user_ids`	`list`	list of user IDs to load. If empty or None, all users are loaded.	`None`
`**kwargs`	`dict`	Additional keyword arguments passed to the `TrajDataFrame` constructor. `encoding` (default `"utf-8"`) is applied when reading each `.plt` file and is not forwarded to `TrajDataFrame`.	`{}`

Returns:

Type	Description
`TrajDataFrame`	a TrajDataFrame with `uid` (string), `lat`/`lng` (float64), and `datetime` (parsed as `timestamp[us]`) columns.

`fastmob.io.file.read(filename, **kwargs)`

Read a trajectory table from disk and return a TrajDataFrame.

CSV and delimited text files are read with pyarrow.csv.read_csv; parquet files are read with pyarrow.parquet.read_table.

Parameters:

Name	Type	Description	Default
`filename`	`str`	path and name of the file to read.	required
`**kwargs`	`dict`	For parquet files, passed to `pyarrow.parquet.read_table`. For CSV/ delimited files, passed to `fastmob.utils._arrow_io.read_delimited` (accepts `delimiter`, `encoding`, `header`, `column_names`, plus any `pyarrow.csv.ConvertOptions` keyword).	`{}`

Returns:

Type	Description
`TrajDataFrame`	object loaded from file.

`fastmob.io.file.write(tdf, filename, **kwargs)`

Write a trajectory dataframe to disk.

CSV and delimited text files are written with pyarrow.csv.write_csv; parquet files are written with pyarrow.parquet.write_table. The input is materialized as a pyarrow.Table via Narwhals first, so any Narwhals-compatible backend (pandas, polars, pyarrow, ...) is accepted regardless of tdf's own backend.

Parameters:

Name	Type	Description	Default
`tdf`	`TrajDataFrame or DataFrame - like`	TrajDataFrame object (or any Narwhals-compatible dataframe) that will be saved.	required
`filename`	`str`	path and name of the output file.	required
`**kwargs`	`dict`	For parquet files, passed to `pyarrow.parquet.write_table`. For CSV/ delimited files, passed to `pyarrow.csv.WriteOptions`.	`{}`

Returns:

Type	Description
`None`