21. Road/Rail Freight

Million tonne-kilometres by Year

Data Source: ITF (2020), https://data.oecd.org/transport/freight-transport.htm

Alternative: “Goods transport”, ITF Transport Statistics (database), https://doi.org/10.1787/g2g5557d-en (accessed on 17 May 2020).

1. Annual data for million tonne-kilometers

import pandas as pd
import missingno as msno

import matplotlib.pyplot as plt
import seaborn as sns;sns.set()

!ls assets

DP_LIVE_17052020151422582.csv		   eurostats_load_19_com.csv
ITF_GOODS_TRANSPORT_17052020153611120.csv  export
ITF_INV-MTN_DATA_17052020174930980.csv	   ocean-freight-infra
airfreight-infra			   road-freight-eurostats

Load Data

df =pd.read_csv('assets/DP_LIVE_17052020151422582.csv')

22. Data Quality

msno.matrix(df)

<AxesSubplot:>

Flag codes indicates:

    1. B: Break
    2. E: Estimated value
    3. P: Provisional value

df["Flag Codes"].unique()

array([nan, 'E', 'B', 'P', 'M'], dtype=object)

df.head()

	LOCATION	INDICATOR	SUBJECT	MEASURE	FREQUENCY	TIME	Value	Flag Codes
0	AUS	FREIGHTTRANSP	RAIL	MLN_TONNEKM	A	1970	36006.0	NaN
1	AUS	FREIGHTTRANSP	RAIL	MLN_TONNEKM	A	1971	39704.0	NaN
2	AUS	FREIGHTTRANSP	RAIL	MLN_TONNEKM	A	1972	42661.0	NaN
3	AUS	FREIGHTTRANSP	RAIL	MLN_TONNEKM	A	1973	46711.0	NaN
4	AUS	FREIGHTTRANSP	RAIL	MLN_TONNEKM	A	1974	54075.0	NaN

Check the columns

df.LOCATION.unique()

array(['AUS', 'AUT', 'BEL', 'CAN', 'CZE', 'DNK', 'FIN', 'FRA', 'DEU',
       'GRC', 'HUN', 'ISL', 'IRL', 'ITA', 'JPN', 'KOR', 'LUX', 'MEX',
       'NLD', 'NZL', 'NOR', 'POL', 'PRT', 'SVK', 'ESP', 'SWE', 'CHE',
       'TUR', 'GBR', 'USA', 'ALB', 'ARM', 'AZE', 'BLR', 'BIH', 'BGR',
       'CHN', 'HRV', 'EST', 'GEO', 'IND', 'LVA', 'LIE', 'LTU', 'MKD',
       'MLT', 'MDA', 'ROU', 'RUS', 'SVN', 'UKR', 'SRB', 'MNE', 'ISR',
       'ARG', 'KAZ', 'CHL', 'ARE'], dtype=object)

df.INDICATOR.unique()

array(['FREIGHTTRANSP'], dtype=object)

df.SUBJECT.unique()

array(['RAIL', 'ROAD', 'INLAND', 'COAST'], dtype=object)

df.MEASURE.unique()

array(['MLN_TONNEKM'], dtype=object)

23. Road Freight Time Series

fig, ax = plt.subplots(figsize=(10,6.18))

df_rf = df.loc[
    (
        df.SUBJECT == 'ROAD'
    ) & (
        df.LOCATION.isin(
            df.groupby('LOCATION').sum().Value.sort_values().tail(10).index
        )
    )
]

sns.lineplot(
    x='TIME',
    y='Value',
    data=df_rf,
    hue='LOCATION',
    marker='o'
)

ax.set_yscale('log')
ax.set_ylabel('Million Tonne-Kilometers')
ax.set_xlabel('Year')

/tmp/ipykernel_2550/3688438137.py:8: FutureWarning: The default value of numeric_only in DataFrameGroupBy.sum is deprecated. In a future version, numeric_only will default to False. Either specify numeric_only or select only columns which should be valid for the function.
  df.groupby('LOCATION').sum().Value.sort_values().tail(10).index

Text(0.5, 0, 'Year')

24. Investments

road_infra_investment_data_path = "assets/ITF_INV-MTN_DATA_17052020174930980.csv"

df_invest = pd.read_csv(road_infra_investment_data_path)

df_invest.head()

	COUNTRY	Country	VARIABLE	Variable	MEASURE	Measure	YEAR	Year	Unit Code	Unit	PowerCode Code	PowerCode	Reference Period Code	Reference Period	Value	Flag Codes	Flags
0	ALB	Albania	I-INV-TOT-INLD	Total inland transport infrastructure investment	NAT	National currency	1996	1996	ALL	Lek	0	Units	NaN	NaN	2.402000e+09	NaN	NaN
1	ALB	Albania	I-INV-TOT-INLD	Total inland transport infrastructure investment	NAT	National currency	1997	1997	ALL	Lek	0	Units	NaN	NaN	2.441000e+09	NaN	NaN
2	ALB	Albania	I-INV-TOT-INLD	Total inland transport infrastructure investment	NAT	National currency	1998	1998	ALL	Lek	0	Units	NaN	NaN	4.765000e+09	NaN	NaN
3	ALB	Albania	I-INV-TOT-INLD	Total inland transport infrastructure investment	NAT	National currency	1999	1999	ALL	Lek	0	Units	NaN	NaN	1.099700e+10	NaN	NaN
4	ALB	Albania	I-INV-TOT-INLD	Total inland transport infrastructure investment	NAT	National currency	2000	2000	ALL	Lek	0	Units	NaN	NaN	1.464200e+10	NaN	NaN

msno.matrix(df_invest)

<AxesSubplot:>

Check data quality

df_invest.VARIABLE.value_counts()

I-INV-RD          3341
I-INV-TOT-INLD    3042
I-INV-MTN-RD      2713
I-INV-RD-MOTW     1732
Name: VARIABLE, dtype: int64

df_invest.MEASURE.value_counts()

NAT          3644
EUR_CONST    3598
EUR          3586
Name: MEASURE, dtype: int64

df_invest_ri = df_invest.loc[
    (
        df_invest.VARIABLE == 'I-INV-RD'
    ) & (
        df_invest.MEASURE == 'EUR'
    ) & (
        df_invest.COUNTRY.isin(
            df_invest.groupby('COUNTRY').sum().Value.sort_values().tail(10).index
        )
    )
]

/tmp/ipykernel_2550/1014484328.py:8: FutureWarning: The default value of numeric_only in DataFrameGroupBy.sum is deprecated. In a future version, numeric_only will default to False. Either specify numeric_only or select only columns which should be valid for the function.
  df_invest.groupby('COUNTRY').sum().Value.sort_values().tail(10).index

fig, ax = plt.subplots(figsize=(10,6.18))

sns.lineplot(
    x='YEAR',
    y='Value',
    data=df_invest_ri,
    hue='COUNTRY',
    marker='o'
)

ax.set_yscale('log')
ax.set_ylabel('Investment in Euros')
ax.set_xlabel('Year')

Text(0.5, 0, 'Year')