disdrodb 0.1.1__py3-none-any.whl → 0.1.3__py3-none-any.whl

disdrodb/l0/readers/LPM/SWITZERLAND/INNERERIZ_LPM.py

@@ -0,0 +1,195 @@
+#!/usr/bin/env python3
+
+# -----------------------------------------------------------------------------.
+# Copyright (c) 2021-2023 DISDRODB developers
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+# -----------------------------------------------------------------------------.
+"""DISDRODB reader for GID LPM sensors not measuring wind."""
+import pandas as pd
+
+from disdrodb.l0.l0_reader import is_documented_by, reader_generic_docstring
+from disdrodb.l0.l0a_processing import read_raw_text_file
+
+
+@is_documented_by(reader_generic_docstring)
+def reader(
+    filepath,
+    logger=None,
+):
+    """Reader."""
+    ##------------------------------------------------------------------------.
+    #### - Define raw data headers
+    column_names = ["TO_BE_SPLITTED"]
+
+    ##------------------------------------------------------------------------.
+    #### Define reader options
+    # - For more info: https://pandas.pydata.org/docs/reference/api/pandas.read_csv.html
+    reader_kwargs = {}
+
+    # - Define delimiter
+    reader_kwargs["delimiter"] = "\\n"
+
+    # - Avoid first column to become df index !!!
+    reader_kwargs["index_col"] = False
+
+    # Since column names are expected to be passed explicitly, header is set to None
+    reader_kwargs["header"] = None
+
+    # - Number of rows to be skipped at the beginning of the file
+    reader_kwargs["skiprows"] = None
+
+    # - Define behaviour when encountering bad lines
+    reader_kwargs["on_bad_lines"] = "skip"
+
+    # - Define reader engine
+    #   - C engine is faster
+    #   - Python engine is more feature-complete
+    reader_kwargs["engine"] = "python"
+
+    # - Define on-the-fly decompression of on-disk data
+    #   - Available: gzip, bz2, zip
+    reader_kwargs["compression"] = "infer"
+
+    # - Strings to recognize as NA/NaN and replace with standard NA flags
+    #   - Already included: '#N/A', '#N/A N/A', '#NA', '-1.#IND', '-1.#QNAN',
+    #                       '-NaN', '-nan', '1.#IND', '1.#QNAN', '<NA>', 'N/A',
+    #                       'NA', 'NULL', 'NaN', 'n/a', 'nan', 'null'
+    reader_kwargs["na_values"] = ["na", "", "error"]
+
+    ##------------------------------------------------------------------------.
+    #### Read the data
+    df = read_raw_text_file(
+        filepath=filepath,
+        column_names=column_names,
+        reader_kwargs=reader_kwargs,
+        logger=logger,
+    )
+
+    ##------------------------------------------------------------------------.
+    #### Adapt the dataframe to adhere to DISDRODB L0 standards
+    # Count number of delimiters to identify valid rows
+    df = df[df["TO_BE_SPLITTED"].str.count(";") == 520]
+
+    # Split by ; delimiter (before raw drop number)
+    df = df["TO_BE_SPLITTED"].str.split(";", expand=True, n=79)
+
+    # Assign column names
+    column_names = [
+        "start_identifier",
+        "device_address",
+        "sensor_serial_number",
+        "sensor_date",
+        "sensor_time",
+        "weather_code_synop_4677_5min",
+        "weather_code_synop_4680_5min",
+        "weather_code_metar_4678_5min",
+        "precipitation_rate_5min",
+        "weather_code_synop_4677",
+        "weather_code_synop_4680",
+        "weather_code_metar_4678",
+        "precipitation_rate",
+        "rainfall_rate",
+        "snowfall_rate",
+        "precipitation_accumulated",
+        "mor_visibility",
+        "reflectivity",
+        "quality_index",
+        "max_hail_diameter",
+        "laser_status",
+        "static_signal",
+        "laser_temperature_analog_status",
+        "laser_temperature_digital_status",
+        "laser_current_analog_status",
+        "laser_current_digital_status",
+        "sensor_voltage_supply_status",
+        "current_heating_pane_transmitter_head_status",
+        "current_heating_pane_receiver_head_status",
+        "temperature_sensor_status",
+        "current_heating_voltage_supply_status",
+        "current_heating_house_status",
+        "current_heating_heads_status",
+        "current_heating_carriers_status",
+        "control_output_laser_power_status",
+        "reserve_status",
+        "temperature_interior",
+        "laser_temperature",
+        "laser_current_average",
+        "control_voltage",
+        "optical_control_voltage_output",
+        "sensor_voltage_supply",
+        "current_heating_pane_transmitter_head",
+        "current_heating_pane_receiver_head",
+        "temperature_ambient",
+        "current_heating_voltage_supply",
+        "current_heating_house",
+        "current_heating_heads",
+        "current_heating_carriers",
+        "number_particles",
+        "number_particles_internal_data",
+        "number_particles_min_speed",
+        "number_particles_min_speed_internal_data",
+        "number_particles_max_speed",
+        "number_particles_max_speed_internal_data",
+        "number_particles_min_diameter",
+        "number_particles_min_diameter_internal_data",
+        "number_particles_no_hydrometeor",
+        "number_particles_no_hydrometeor_internal_data",
+        "number_particles_unknown_classification",
+        "number_particles_unknown_classification_internal_data",
+        "number_particles_class_1",
+        "number_particles_class_1_internal_data",
+        "number_particles_class_2",
+        "number_particles_class_2_internal_data",
+        "number_particles_class_3",
+        "number_particles_class_3_internal_data",
+        "number_particles_class_4",
+        "number_particles_class_4_internal_data",
+        "number_particles_class_5",
+        "number_particles_class_5_internal_data",
+        "number_particles_class_6",
+        "number_particles_class_6_internal_data",
+        "number_particles_class_7",
+        "number_particles_class_7_internal_data",
+        "number_particles_class_8",
+        "number_particles_class_8_internal_data",
+        "number_particles_class_9",
+        "number_particles_class_9_internal_data",
+        "raw_drop_number",
+    ]
+    df.columns = column_names
+
+    # Remove checksum from raw_drop_number
+    df["raw_drop_number"] = df["raw_drop_number"].str.rsplit(";", n=2, expand=True)[0]
+
+    # Define datetime "time" column
+    df["time"] = df["sensor_date"] + "-" + df["sensor_time"]
+    df["time"] = pd.to_datetime(df["time"], format="%d.%m.%y-%H:%M:%S", errors="coerce")
+
+    # Drop row if start_identifier different than 00
+    df = df[df["start_identifier"].astype(str) == "00"]
+
+    # Drop rows with invalid raw_drop_number
+    df = df[df["raw_drop_number"].astype(str).str.len() == 1759]
+
+    # Drop columns not agreeing with DISDRODB L0 standards
+    columns_to_drop = [
+        "start_identifier",
+        "device_address",
+        "sensor_serial_number",
+        "sensor_date",
+        "sensor_time",
+    ]
+    df = df.drop(columns=columns_to_drop)
+    return df

disdrodb/l0/readers/PARSIVEL/GPM/PIERS.py

@@ -0,0 +1,105 @@
+#!/usr/bin/env python3
+# -----------------------------------------------------------------------------.
+# Copyright (c) 2021-2023 DISDRODB developers
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+# -----------------------------------------------------------------------------.
+import pandas as pd
+
+from disdrodb.l0.l0_reader import is_documented_by, reader_generic_docstring
+from disdrodb.l0.l0a_processing import read_raw_text_file
+
+
+@is_documented_by(reader_generic_docstring)
+def reader(
+    filepath,
+    logger=None,
+):
+    """Reader."""
+    ##------------------------------------------------------------------------.
+    #### Define column names
+    column_names = ["time", "TO_BE_SPLITTED"]
+
+    ##------------------------------------------------------------------------.
+    #### Define reader options
+    reader_kwargs = {}
+    # - Define delimiter
+    reader_kwargs["delimiter"] = ";"
+    # - Skip first row as columns names
+    reader_kwargs["header"] = None
+    reader_kwargs["skiprows"] = 0
+    # - Skip file with encoding errors
+    reader_kwargs["encoding_errors"] = "ignore"
+    # - Avoid first column to become df index !!!
+    reader_kwargs["index_col"] = False
+    # - Define behaviour when encountering bad lines
+    reader_kwargs["on_bad_lines"] = "skip"
+    # - Define reader engine
+    #   - C engine is faster
+    #   - Python engine is more feature-complete
+    reader_kwargs["engine"] = "python"
+    # - Define on-the-fly decompression of on-disk data
+    #   - Available: gzip, bz2, zip
+    reader_kwargs["compression"] = "infer"
+    # - Strings to recognize as NA/NaN and replace with standard NA flags
+    #   - Already included: '#N/A', '#N/A N/A', '#NA', '-1.#IND', '-1.#QNAN',
+    #                       '-NaN', '-nan', '1.#IND', '1.#QNAN', '<NA>', 'N/A',
+    #                       'NA', 'NULL', 'NaN', 'n/a', 'nan', 'null'
+    reader_kwargs["na_values"] = ["na", "", "error", "NA", "-.-"]
+
+    ##------------------------------------------------------------------------.
+    #### Read the data
+    df = read_raw_text_file(
+        filepath=filepath,
+        column_names=column_names,
+        reader_kwargs=reader_kwargs,
+        logger=logger,
+    )
+
+    ##------------------------------------------------------------------------.
+    #### Adapt the dataframe to adhere to DISDRODB L0 standards
+    # Convert time column to datetime
+    df_time = pd.to_datetime(df["time"], format="%Y%m%d%H%M%S", errors="coerce")
+
+    # Split the 'TO_BE_SPLITTED' column
+    df = df["TO_BE_SPLITTED"].str.split(",", expand=True, n=9)
+
+    # Assign column names
+    columns_names = [
+        "station_name",
+        "sensor_status",
+        "sensor_temperature",
+        "number_particles",
+        "rainfall_rate_32bit",
+        "reflectivity_16bit",
+        "mor_visibility",
+        "weather_code_synop_4680",
+        "weather_code_synop_4677",
+        "raw_drop_number",
+    ]
+    df.columns = columns_names
+
+    # Add the time column
+    df["time"] = df_time
+
+    # Drop columns not agreeing with DISDRODB L0 standards
+    df = df.drop(columns=["station_name"])
+
+    # Drop rows with invalid values
+    # --> Ensure that weather_code_synop_4677 has length 2
+    # --> If a previous column is missing it will have 000
+    df = df[df["weather_code_synop_4677"].str.len() == 2]
+
+    # Return the dataframe adhering to DISDRODB L0 standards
+    return df

disdrodb/l0/readers/PARSIVEL/JAPAN/JMA.py

@@ -0,0 +1,128 @@
+#!/usr/bin/env python3
+# -----------------------------------------------------------------------------.
+# Copyright (c) 2021-2023 DISDRODB developers
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+# -----------------------------------------------------------------------------.
+import pandas as pd
+
+from disdrodb.l0.l0_reader import is_documented_by, reader_generic_docstring
+from disdrodb.l0.l0a_processing import read_raw_text_file
+
+
+@is_documented_by(reader_generic_docstring)
+def reader(
+    filepath,
+    logger=None,
+):
+    """Reader."""
+    ##------------------------------------------------------------------------.
+    #### Define column names
+    column_names = ["TO_SPLIT"]
+
+    ##------------------------------------------------------------------------.
+    #### Define reader options
+    reader_kwargs = {}
+
+    # - Define delimiter
+    reader_kwargs["delimiter"] = "\\n"
+
+    # - Skip first row as columns names
+    reader_kwargs["header"] = None
+
+    # - Skip header
+    reader_kwargs["skiprows"] = 0
+
+    # - Define encoding
+    reader_kwargs["encoding"] = "ISO-8859-1"
+
+    # - Avoid first column to become df index !!!
+    reader_kwargs["index_col"] = False
+
+    # - Define behaviour when encountering bad lines
+    reader_kwargs["on_bad_lines"] = "skip"
+
+    # - Define reader engine
+    #   - C engine is faster
+    #   - Python engine is more feature-complete
+    reader_kwargs["engine"] = "python"
+
+    # - Define on-the-fly decompression of on-disk data
+    #   - Available: gzip, bz2, zip
+    # reader_kwargs['compression'] = 'xz'
+
+    # - Strings to recognize as NA/NaN and replace with standard NA flags
+    #   - Already included: '#N/A', '#N/A N/A', '#NA', '-1.#IND', '-1.#QNAN',
+    #                       '-NaN', '-nan', '1.#IND', '1.#QNAN', '<NA>', 'N/A',
+    #                       'NA', 'NULL', 'NaN', 'n/a', 'nan', 'null'
+    reader_kwargs["na_values"] = ["na", "error", "-.-", " NA"]
+
+    ##------------------------------------------------------------------------.
+    #### Read the data
+    df = read_raw_text_file(
+        filepath=filepath,
+        column_names=column_names,
+        reader_kwargs=reader_kwargs,
+        logger=logger,
+    )
+
+    ##------------------------------------------------------------------------.
+    #### Adapt the dataframe to adhere to DISDRODB L0 standards
+    # Remove rows with less than 97 characters (empty spectrum --> 97 characters)
+    df = df[df["TO_SPLIT"].str.len() >= 97]
+
+    # Split into columns and assign name
+    df = df["TO_SPLIT"].str.split(";", expand=True, n=14)
+    columns = [
+        "date",
+        "time",
+        "rainfall_rate_32bit",
+        "rainfall_accumulated_32bit",
+        "weather_code_synop_4680",
+        "weather_code_metar_4678",
+        "weather_code_nws",
+        "reflectivity_32bit",
+        "mor_visibility",
+        "laser_amplitude",
+        "number_particles",
+        "sensor_temperature",
+        "sensor_heating_current",
+        "sensor_battery_voltage",
+        "raw_drop_number",
+    ]
+    df.columns = columns
+
+    # Add datetime time column
+    df["time"] = df["date"] + "-" + df["time"]
+    df["time"] = pd.to_datetime(df["time"], format="%Y/%m/%d-%H:%M:%S", errors="coerce")
+    df = df.drop(columns=["date"])
+
+    # Convert timezone from JST to UTC
+    df = df.set_index("time").tz_localize("Asia/Tokyo").tz_convert(None).reset_index()
+
+    # Preprocess the raw spectrum
+    # - The '<SPECTRUM>ZERO</SPECTRUM>' indicates no drops detected
+    # --> "" generates an array of zeros in L0B processing
+    df["raw_drop_number"] = df["raw_drop_number"].str.replace("<SPECTRUM>ZERO</SPECTRUM>", "")
+
+    # Remove <SPECTRUM> and </SPECTRUM> prefix and suffix from the raw_drop_number field
+    df["raw_drop_number"] = df["raw_drop_number"].str.replace("<SPECTRUM>", "")
+    df["raw_drop_number"] = df["raw_drop_number"].str.replace("</SPECTRUM>", "")
+
+    # Add 0 before every , if , not preceded by a digit
+    # Example: ',,1,,' --> '0,0,1,0,'
+    df["raw_drop_number"] = df["raw_drop_number"].str.replace(r"(?<!\d);", "0;", regex=True)
+
+    # Return the dataframe adhering to DISDRODB L0 standards
+    return df

disdrodb/l0/readers/PARSIVEL/NCAR/PECAN_MOBILE.py

@@ -113,7 +113,7 @@ def reader(
     # --> "" generates an array of zeros in L0B processing
     df["raw_drop_number"] = df["raw_drop_number"].str.replace("<SPECTRUM>ZERO</SPECTRUM>", "")
 
-    # Remove <SPECTRUM> and </SPECTRUM>" acronyms from the raw_drop_number field
+    # Remove <SPECTRUM> and </SPECTRUM> prefix and suffix from the raw_drop_number field
     df["raw_drop_number"] = df["raw_drop_number"].str.replace("<SPECTRUM>", "")
     df["raw_drop_number"] = df["raw_drop_number"].str.replace("</SPECTRUM>", "")
 

disdrodb/l0/readers/PARSIVEL/NCAR/VORTEX2_2009.py

@@ -102,7 +102,7 @@ def reader(
     # --> "" generates an array of zeros in L0B processing
     df["raw_drop_number"] = df["raw_drop_number"].str.replace("<SPECTRUM>ZERO</SPECTRUM>", "")
 
-    # Remove <SPECTRUM> and </SPECTRUM>" acronyms from the raw_drop_number field
+    # Remove <SPECTRUM> and </SPECTRUM> prefix and suffix from the raw_drop_number field
     df["raw_drop_number"] = df["raw_drop_number"].str.replace("<SPECTRUM>", "")
     df["raw_drop_number"] = df["raw_drop_number"].str.replace("</SPECTRUM>", "")
 

disdrodb/l0/readers/PARSIVEL2/BELGIUM/ILVO.py

@@ -0,0 +1,168 @@
+#!/usr/bin/env python3
+
+# -----------------------------------------------------------------------------.
+# Copyright (c) 2021-2023 DISDRODB developers
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+# -----------------------------------------------------------------------------.
+"""DISDRODB reader for UGENT ILVO Parsivel2 raw text data."""
+import tarfile
+
+import numpy as np
+import pandas as pd
+
+from disdrodb.l0.l0_reader import is_documented_by, reader_generic_docstring
+from disdrodb.l0.l0a_processing import read_raw_text_file
+from disdrodb.utils.logger import log_error
+
+
+@is_documented_by(reader_generic_docstring)
+def reader(
+    filepath,
+    logger=None,
+):
+    """Reader."""
+
+    ##------------------------------------------------------------------------.
+    #### Define function to read each txt file inside each daily zip file
+    def read_txt_file(file, filename):
+        """Parse a single txt file within the daily zip file."""
+        ##------------------------------------------------------------------------.
+        #### Define column names
+        column_names = ["TO_PARSE"]
+
+        ##------------------------------------------------------------------------.
+        #### Define reader options
+        reader_kwargs = {}
+        # - Define delimiter
+        reader_kwargs["delimiter"] = "\\n"
+        # - Skip first row as columns names
+        # - Define encoding
+        reader_kwargs["encoding"] = "latin"  # "ISO-8859-1"
+        # - Avoid first column to become df index !!!
+        reader_kwargs["index_col"] = False
+        # - Define behaviour when encountering bad lines
+        reader_kwargs["on_bad_lines"] = "skip"
+        # - Define reader engine
+        #   - C engine is faster
+        #   - Python engine is more feature-complete
+        reader_kwargs["engine"] = "python"
+        # - Define on-the-fly decompression of on-disk data
+        #   - Available: gzip, bz2, zip
+        reader_kwargs["compression"] = "infer"
+        # - Strings to recognize as NA/NaN and replace with standard NA flags
+        #   - Already included: '#N/A', '#N/A N/A', '#NA', '-1.#IND', '-1.#QNAN',
+        #                       '-NaN', '-nan', '1.#IND', '1.#QNAN', '<NA>', 'N/A',
+        #                       'NA', 'NULL', 'NaN', 'n/a', 'nan', 'null'
+        reader_kwargs["na_values"] = ["na", "", "error"]
+
+        ##------------------------------------------------------------------------.
+        #### Read the data
+        df = read_raw_text_file(
+            filepath=file,
+            column_names=column_names,
+            reader_kwargs=reader_kwargs,
+            logger=logger,
+        )
+
+        ##------------------------------------------------------------------------.
+        #### Adapt the dataframe to adhere to DISDRODB L0 standards
+        # Create ID and Value columns
+        df = df["TO_PARSE"].str.split(":", expand=True, n=1)
+        df.columns = ["ID", "Value"]
+
+        # Select only rows with values
+        df = df[df["Value"].apply(lambda x: x is not None)]
+
+        # Drop rows with invalid IDs
+        valid_id_str = np.char.rjust(np.arange(0, 94).astype(str), width=2, fillchar="0")
+        df = df[df["ID"].astype(str).isin(valid_id_str)]
+
+        # Reshape dataframe
+        df = df.set_index("ID").T
+
+        # Assign column names
+        column_dict = {
+            "01": "rainfall_rate_32bit",
+            "02": "rainfall_accumulated_32bit",
+            "03": "weather_code_synop_4680",
+            "04": "weather_code_synop_4677",
+            "05": "weather_code_metar_4678",
+            "06": "weather_code_nws",
+            "07": "reflectivity_32bit",
+            "08": "mor_visibility",
+            "09": "sample_interval",
+            "10": "laser_amplitude",
+            "11": "number_particles",
+            "12": "sensor_temperature",
+            # "13": "sensor_serial_number",
+            # "14": "firmware_iop",
+            # "15": "firmware_dsp",
+            "16": "sensor_heating_current",
+            "17": "sensor_battery_voltage",
+            "18": "sensor_status",
+            # "19": "start_time",
+            # "20": "sensor_time",
+            # "21": "sensor_date",
+            # "22": "station_name",
+            # "23": "station_number",
+            "24": "rainfall_amount_absolute_32bit",
+            "25": "error_code",
+            "26": "sensor_temperature_pcb",
+            "27": "sensor_temperature_receiver",
+            "28": "sensor_temperature_trasmitter",
+            "30": "rainfall_rate_16_bit_30",
+            "31": "rainfall_rate_16_bit_1200",
+            "32": "rainfall_accumulated_16bit",
+            "34": "rain_kinetic_energy",
+            "35": "snowfall_rate",
+            "90": "raw_drop_concentration",
+            "91": "raw_drop_average_velocity",
+            "93": "raw_drop_number",
+        }
+
+        df = df.rename(column_dict, axis=1)
+
+        # Keep only columns defined in the dictionary
+        df = df[list(column_dict.values())]
+
+        # Define datetime "time" column from filename
+        datetime_str = filename.replace(".txt", "").split("_")[1]
+        df["time"] = pd.to_datetime(datetime_str, format="%Y%m%d%H%M%S")
+
+        return df
+
+    # ---------------------------------------------------------------------.
+    #### Iterate over all files (aka timesteps) in the daily zip archive
+    # - Each file contain a single timestep !
+    list_df = []
+    with tarfile.open(filepath, "r:gz") as tar:
+        members = sorted(tar.getmembers(), key=lambda m: m.name)
+        for member in members:
+            filename = member.name
+            if member.isfile() and filename.endswith(".txt"):
+                # Open file
+                with tar.extractfile(member) as file:
+                    try:
+                        df = read_txt_file(file=file, filename=filename)
+                        list_df.append(df)
+                    except Exception as e:
+                        msg = f"An error occurred while reading {filename}. The error is: {e}."
+                        log_error(logger=logger, msg=msg, verbose=True)
+
+    # Concatenate all dataframes into a single one
+    df = pd.concat(list_df)
+
+    # ---------------------------------------------------------------------.
+    return df