Skip to content

Converters

influpaint.utils.converters

dataframe_to_xarray(df, season_setup=None, xarray_name='data', xarrax_features='value', date_column='week_enddate', value_column='value', pad=True)

Convert a long form dataframe to an xarray. Dataframe must have columns: - location_code - value The dataset is a xarray object stored as netcdf on disk. It has dimensions (feature, date, place) where date and place are padded to have dimension 64. - dates are Saturdays - places are location from Flusight data locations - samples are integers

Source code in influpaint/utils/converters.py 
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
def dataframe_to_xarray(
    df: pd.DataFrame,
    season_setup: SeasonAxis = None,
    xarray_name="data",
    xarrax_features="value",
    date_column="week_enddate",
    value_column="value",
    pad=True,
) -> xr.DataArray:
    """
    Convert a long form dataframe to an xarray. Dataframe must have columns:
    - location_code
    - value
    The dataset is a xarray object stored as netcdf on disk. 
    It has dimensions `(feature, date, place)` 
    where date and place are padded to have dimension 64.
    - dates are Saturdays
    - places are location from Flusight data locations
    - samples are integers
    """

    df_piv = df.reset_index(drop=False).pivot(columns="location_code", values=value_column, index=date_column)

    if not isinstance(xarrax_features, list):
        xarrax_features = [xarrax_features]

    if season_setup is None:
        print(" ⚠️ No season_setup provided, using all locations in the dataframe.")
        places = df_piv.columns.to_list()
    else:
        df_piv = df_piv[
            season_setup.locations_df["location_code"]
        ]  # make sure order is right w.r.t flusight_locations
        places = season_setup.locations_df["location_code"]
        df_piv = df_piv.sort_index(axis=1)

    df_xarr = xr.DataArray(
        np.array([df_piv.to_numpy()]),
        name=xarray_name,
        coords={
            "feature": xarrax_features,
            "date": list(df_piv.index),
            "place": places,
        },
        dims=["feature", "date", "place"],
    )
    if pad:
        df_xarr = df_xarr.pad(
            {
                "date": (0, 64 - len(df_xarr.date)),
                "place": (0, 64 - len(df_xarr.place)),
            },
            mode="constant",
            constant_values=0,
        )

    return df_xarr