Export data from a pvcaptest CapData object to a PVsyst standard format csv file

to_pvsyst_csv.py

def to_pvsyst_csv(
    path,
    cd_obj,
    start_date,
    loc,
    fname_prefix,
    end_date=None,
    ghi_col_name=None,
    dhi_col_name=None,
    poa=False,
    export=True,
    filtered=True,
    return_ix=True,
):
    """
    Given a CapData object export a PVsyst csv format weather file.

    Function does not resample!  PVsyst requires hourly intervals
    for simulation.  To succesfully import the output of this
    function into PVsyst you must pass a CapData object where the
    index of the dataframe attribute has an hourly frequency.

    Output file has GHI, ambient temperature, and wind speed data
    and expects to find this data in the CapData object identified
    by the the tranlsation dictionary and the regression translation
    dictionary.

    Adds units assuming that the GHI data is W/m2, the temperature
    is in Celsius, and the wind speed is in m/sec.

    Any missing data is filled with zeros.

    path : str
        Path to export csv file.
    cd_obj : CapData
        captest CapData object.
    start_date : str
    loc : dict
        Location dictionary.  Same as that used in the Captest csky function, but with
        the city and state added as strings. Below is an example loc dictionary.

        loc = {'latitude': 30.275128,
               'longitude': -97.740350
               'altitude': 300,
               'tz': 'America/Chicago',
               'city': 'Austin',
               'state': 'TX'}

    fname_prefix : str
        Prefix for the filename of the exported file.  Also used in header data of
        exported file as the site name.
    end_date : str, default None
        End date of data to be exported.  By default uses all data from the passed
        start date to the end of the data available.
    ghi_col_name : str, default None
        A column name used to select the column with GHI data for inclusion in the
        exported file. By default an attempt is made to identify the GHI column
        from the translation dictionary.
    dhi_col_name : str, default None
        A column name used to select the column with DHI data for inclusion in the
        exported file.
    poa : bool, default False
        Set to true to include POA irradiance in the data exported.
    export : bool, default True
        If true exports the data, if false will return a tuple. The first value
        is a pvsyst format dataframe and the second value is the header data as
        a string.
    filtered : bool, default True
        By default exports data from the data_filtered attribute.  Set to False to use
        the data in the df attribute of the passed CapData object.
    return_ix : bool, default True
        By default returns the index of the exported dataframe.
    """
    if filtered:
        df = cd_obj.data_filtered.copy()
    else:
        df = cd_obj.data.copy()

    # select portion of data to output and fill missing timestamps with zeros
    if end_date is None:
        end_date = df.index[-1]
        df = df.loc[start_date:end_date, :]
    else:
        df = df.loc[start_date:end_date, :]

    complete_index = pd.date_range(start=start_date, end=end_date, freq="60min")
    df = df.reindex(index=complete_index, fill_value=0)

    # initialize empty dataframe to hold data and return from function
    df_pvsyst_format = pd.DataFrame()

    # add date and time columns to dataframe
    df_pvsyst_format["Year"] = df.index.year.astype("int")
    df_pvsyst_format["Month"] = df.index.month.astype("int")
    df_pvsyst_format["Day"] = df.index.day.astype("int")
    df_pvsyst_format["Hour"] = df.index.hour.astype("int")

    # can't add the below data w/o the index
    df_pvsyst_format.index = df.index

    # try to find a ghi column
    if ghi_col_name is None:
        ghi_trans_keys = [key for key in cd_obj.trans_keys if "ghi" in key]

        if len(ghi_trans_keys) == 1:
            ghi_col_name = cd_obj.column_groups[ghi_trans_keys[0]]
        elif len(ghi_trans_keys) == 0:
            print("There are no GHI columns identified in the translation dictionary.")
            return
        else:
            print(
                "There is more than one ghi column_groupslation key.  Pass one to ghi_col_name:"
            )
            for ghi_key in ghi_column_groups_keys:
                print(ghi_key)
            return

    # add GHI, ambient temp, and wind speed to output dataframe
    df_pvsyst_format["GHI"] = df[ghi_col_name]
    if dhi_col_name is not None:
        try:
            df_pvsyst_format["DHI"] = df[dhi_col_name]
        except KeyError:
            warnings.warn("DHI column not found in data.")
    if poa:
        df_pvsyst_format["GPI"] = cd_obj.get_reg_cols("poa")
    df_pvsyst_format["Tamb"] = cd_obj.get_reg_cols("t_amb")
    df_pvsyst_format["WindVel"] = cd_obj.get_reg_cols("w_vel")

    units = {
        "Year": "",
        "Month": "",
        "Day": "",
        "Hour": "",
        "GHI": "W/m2",
        "DHI": "W/m2",
        "GPI": "W/m2",
        "Tamb": "deg.C",
        "WindVel": "m/sec",
    }

    def add_units(df, units):
        df.columns = pd.MultiIndex.from_arrays(
            [
                df.columns,
                [units[col] for col in df.columns],
            ]
        )
        return df

    output_columns = [
        "Year",
        "Month",
        "Day",
        "Hour",
        "GHI",
        "DHI",
        "GPI",
        "Tamb",
        "WindVel",
    ]

    df_pvsyst_format = add_units(df_pvsyst_format, units).reindex(
        columns=output_columns, level=0
    )

    # Replace missing data (NaNs) with zeros
    df_pvsyst_format.fillna(value=0, inplace=True)

    header_str = build_pvsyst_header_str(
        df_pvsyst_format, fname_prefix, loc, country="USA"
    )

    # filename of exported file
    pvsyst_export_filename = (
        fname_prefix
        + "_"
        + df_pvsyst_format.index[0].strftime("%Y%m%d")
        + "to"
        + df_pvsyst_format.index[-1].strftime("%Y%m%d")
    )

    # create csv file
    if export:
        # output data
        df_pvsyst_format.to_csv(path + pvsyst_export_filename + ".csv", index=False)
        # prepend header rows to exported file
        line_prepender(path + pvsyst_export_filename + ".csv", header_str)
        print("Exported file is called: {}.csv".format(pvsyst_export_filename))
        return df_pvsyst_format.index
    else:
        return df_pvsyst_format, header_str