Skip to content

Historical

HistoricalTrips

Source code in lyft_bikes/historical/historical.py 
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
class HistoricalTrips:
    def __init__(self, dates: Dates, downloader: Downloader):
        self.dates = dates
        self.downloader = downloader

    def get_months_trips(self, year: int, month: int) -> pd.DataFrame:
        """Return pandas.DataFrame for a given year and month"""
        date = self.dates.to_date(year, month)
        self.dates.check_valid(date)

        return self.downloader.read(date=date)

    def read(
        self,
        start_date: str,
        start_time_col: str = "started_at",
        end_date: Union[str, None] = None,
    ) -> pd.DataFrame:
        """Return historical trips for a given range of dates

        Args:
            start_date: start date for the data in %Y-%m-%d format
            start_col_col: column name for the start time
            end_date: end date in the same format. Defaults to last date available

        Returns:
            Historical trip DataFrame for the date range provided.

        Examples:
            Read trips from Jan 1st 2021 until Feb 1st 2021

            >>> trips = HistoricalTrips()
            >>> df_trips = trips.read(start_date="2021-01-01", end_date="2021-02-01")

            Read trips from Jan 1st 2021 until the last date available

            >>> df_trips = trips.read(start_date="2021-01-01")

        """
        if end_date is None:
            end_date = str(self.dates.last_date)

        df_trips = pd.concat(
            [
                self.get_months_trips(date.year, date.month)
                for date in self.dates.create_date_range(start_date, end_date)
            ],
            ignore_index=True,
        )

        try:
            date = pd.to_datetime(df_trips[start_time_col]).dt.date.astype(str)
            idx = (date >= start_date) & (date <= end_date)
            return df_trips.loc[idx, :].reset_index(drop=True)
        except KeyError:
            warnings.warn(
                f"The column {start_time_col} does not exist. Returning all trips instead.",
                UserWarning,
            )
            return df_trips

get_months_trips(year, month)

Return pandas.DataFrame for a given year and month

Source code in lyft_bikes/historical/historical.py 
33
34
35
36
37
38
def get_months_trips(self, year: int, month: int) -> pd.DataFrame:
    """Return pandas.DataFrame for a given year and month"""
    date = self.dates.to_date(year, month)
    self.dates.check_valid(date)

    return self.downloader.read(date=date)

read(start_date, start_time_col='started_at', end_date=None)

Return historical trips for a given range of dates

Parameters:

Name Type Description Default
start_date str

start date for the data in %Y-%m-%d format

 required
start_col_col

column name for the start time

 required
end_date Union[str, None]

end date in the same format. Defaults to last date available

 None

Returns:

Type Description
pd.DataFrame

Historical trip DataFrame for the date range provided.

Examples:

Read trips from Jan 1st 2021 until Feb 1st 2021

>>> trips = HistoricalTrips()
>>> df_trips = trips.read(start_date="2021-01-01", end_date="2021-02-01")

Read trips from Jan 1st 2021 until the last date available

>>> df_trips = trips.read(start_date="2021-01-01")
Source code in lyft_bikes/historical/historical.py 
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
def read(
    self,
    start_date: str,
    start_time_col: str = "started_at",
    end_date: Union[str, None] = None,
) -> pd.DataFrame:
    """Return historical trips for a given range of dates

    Args:
        start_date: start date for the data in %Y-%m-%d format
        start_col_col: column name for the start time
        end_date: end date in the same format. Defaults to last date available

    Returns:
        Historical trip DataFrame for the date range provided.

    Examples:
        Read trips from Jan 1st 2021 until Feb 1st 2021

        >>> trips = HistoricalTrips()
        >>> df_trips = trips.read(start_date="2021-01-01", end_date="2021-02-01")

        Read trips from Jan 1st 2021 until the last date available

        >>> df_trips = trips.read(start_date="2021-01-01")

    """
    if end_date is None:
        end_date = str(self.dates.last_date)

    df_trips = pd.concat(
        [
            self.get_months_trips(date.year, date.month)
            for date in self.dates.create_date_range(start_date, end_date)
        ],
        ignore_index=True,
    )

    try:
        date = pd.to_datetime(df_trips[start_time_col]).dt.date.astype(str)
        idx = (date >= start_date) & (date <= end_date)
        return df_trips.loc[idx, :].reset_index(drop=True)
    except KeyError:
        warnings.warn(
            f"The column {start_time_col} does not exist. Returning all trips instead.",
            UserWarning,
        )
        return df_trips