Pandas

"Bai yun giant panda" von Mfield, Matthew Field, http://www.photography.mattfield.com - Eigenes Werk. Lizenziert unter CC BY-SA 3.0 über Wikimedia Commons.

Python Data Analysis Library¶

pandas is an open source, BSD-licensed library providing high-performance, easy-to-use data structures and data analysis tools for the Python programming language.

http://pandas.pydata.org/

About this talk¶

Introduction to Pandas
I will show all code
walk through the basics quickly
and see some of the fun stuff!
slides online at http://trescher.fr/pandas/

About me¶

PhD Student in Physics at
interested in Python since 2005:
- Web
- Numerics / Data
- Plotting

Preparation¶

In [1]:

from __future__ import print_function, division

from matplotlib import pyplot as plt
import numpy as np

%matplotlib inline

In [2]:

import pandas as pd

In [3]:

pd.set_option('display.max_rows', 8)

Basics

Loading Data¶

Data from DWD:
ftp://ftp-cdc.dwd.de/pub/CDC/observations_germany/climate/daily/kl/historical/

Daily data, they also have hourly / monthly data

Station IDs:

Berlin (Dahlem - FU) 403

Freiburg 1443

Disclaimer:
This is only an example, I'm not an expert in meteorology.

In [4]:

df1 = pd.read_csv('produkt_klima_Tageswerte_19500101_20141231_00403.txt', 
                  sep=";", 
                  index_col=1, 
                  parse_dates=[1], 
                  skipinitialspace=True)
df2 = pd.read_csv('produkt_klima_Tageswerte_19490501_20141231_01443.txt', 
                  sep=";", 
                  index_col=1, 
                  parse_dates=[1], 
                  skipinitialspace=True)

#last row invalid for obscure reasons
df1.drop(df1.index[-1], inplace=True)
df2.drop(df2.index[-1], inplace=True)

There is also:

read excel
read hdf
read sql
...

Look at Data¶

In [5]:

df1

Out[5]:

	STATIONS_ID	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE	eor
MESS_DATUM
1950-01-01	403	5	-3.2	4.0	5.0	1025.60	83.00	-999	-1.1	-4.9	-6.3	-999	2.2	7	-999.000	0	eor
1950-01-02	403	5	1.0	6.1	8.0	1005.60	95.00	-999	2.2	-3.7	-5.3	-999	12.6	8	-999.000	0	eor
1950-01-03	403	5	2.8	6.5	5.0	996.60	86.00	-999	3.9	1.7	-1.4	-999	0.5	1	-999.000	0	eor
1950-01-04	403	5	-0.1	5.2	7.7	999.50	85.00	-999	2.1	-0.9	-2.3	-999	0.5	7	-999.000	0	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2014-12-28	403	3	-5.1	3.8	4.6	1012.22	90.21	-999	-2.3	-7.6	-10.1	-999	3.6	7	1.333	0	eor
2014-12-29	403	3	-3.6	4.5	5.6	1018.05	95.92	-999	-1.9	-6.0	-13.8	-999	0.0	7	6.117	11	eor
2014-12-30	403	3	-1.8	5.2	7.9	1020.31	97.42	-999	1.1	-3.5	-4.6	-999	0.6	8	0.000	8	eor
2014-12-31	403	3	2.8	7.4	8.0	1019.90	99.00	-999	4.9	1.1	0.0	-999	0.9	6	0.000	6	eor

23741 rows × 17 columns

Indexing¶

http://pandas.pydata.org/pandas-docs/stable/indexing.html

In [6]:

df1.columns

Out[6]:

Index([u'STATIONS_ID', u'QUALITAETS_NIVEAU', u'LUFTTEMPERATUR', u'DAMPFDRUCK', u'BEDECKUNGSGRAD', u'LUFTDRUCK_STATIONSHOEHE', u'REL_FEUCHTE', u'WINDGESCHWINDIGKEIT', u'LUFTTEMPERATUR_MAXIMUM', u'LUFTTEMPERATUR_MINIMUM', u'LUFTTEMP_AM_ERDB_MINIMUM', u'WINDSPITZE_MAXIMUM', u'NIEDERSCHLAGSHOEHE', u'NIEDERSCHLAGSHOEHE_IND', u'SONNENSCHEINDAUER', u'SCHNEEHOEHE', u'eor'], dtype='object')

In [7]:

df1["LUFTTEMPERATUR"]

Out[7]:

MESS_DATUM
1950-01-01   -3.2
1950-01-02    1.0
...
2014-12-30   -1.8
2014-12-31    2.8
Name: LUFTTEMPERATUR, Length: 23741

it's called a "Series"

Label based

In [8]:

df1.loc["1970-01-01"]["LUFTTEMPERATUR"]

Out[8]:

-12.800000000000001

and integer-index based

In [9]:

df1.iloc[120:123]

Out[9]:

	STATIONS_ID	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE	eor
MESS_DATUM
1950-05-01	403	5	16.1	9.5	0.7	1017.9	55	-999	21.5	5.6	1.0	-999	0.0	0	12.8	0	eor
1950-05-02	403	5	17.4	10.0	1.3	1011.6	53	-999	23.9	10.5	6.7	-999	0.2	1	12.6	0	eor
1950-05-03	403	5	11.2	11.6	6.7	1011.0	85	-999	16.7	9.8	9.8	-999	1.1	1	0.2	0	eor

And a little bit of magic¶

In [10]:

df1["1970"]

Out[10]:

	STATIONS_ID	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE	eor
MESS_DATUM
1970-01-01	403	5	-12.8	2.0	8.0	997.6	84	-999	-11.4	-15.6	-15.6	-999	0.6	7	0.0	8	eor
1970-01-02	403	5	-4.9	3.9	8.0	990.6	92	-999	-2.8	-12.7	-12.6	-999	6.6	7	0.0	9	eor
1970-01-03	403	5	-1.8	5.2	7.0	989.9	96	-999	0.2	-4.0	-10.1	-999	3.2	7	0.0	19	eor
1970-01-04	403	5	-5.9	3.7	1.0	993.8	88	-999	-1.2	-8.0	-12.6	-999	0.0	0	6.4	23	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
1970-12-28	403	5	-2.4	4.4	7.7	997.5	88	-999	-2.1	-4.1	-4.6	-999	7.4	7	0.0	2	eor
1970-12-29	403	5	-2.2	4.8	8.0	997.2	91	-999	-1.5	-2.5	-3.0	-999	0.6	7	0.0	12	eor
1970-12-30	403	5	-2.1	4.8	8.0	988.7	91	-999	-1.1	-2.7	-3.1	-999	1.2	7	0.0	9	eor
1970-12-31	403	5	-7.8	3.1	5.3	999.5	89	-999	-2.3	-9.6	-12.2	-999	0.0	7	0.2	10	eor

365 rows × 17 columns

In [11]:

df1["1970":"1980"]

Out[11]:

	STATIONS_ID	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE	eor
MESS_DATUM
1970-01-01	403	5	-12.8	2.0	8.0	997.6	84	-999	-11.4	-15.6	-15.6	-999	0.6	7	0.0	8	eor
1970-01-02	403	5	-4.9	3.9	8.0	990.6	92	-999	-2.8	-12.7	-12.6	-999	6.6	7	0.0	9	eor
1970-01-03	403	5	-1.8	5.2	7.0	989.9	96	-999	0.2	-4.0	-10.1	-999	3.2	7	0.0	19	eor
1970-01-04	403	5	-5.9	3.7	1.0	993.8	88	-999	-1.2	-8.0	-12.6	-999	0.0	0	6.4	23	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
1980-12-28	403	5	1.0	5.4	7.7	1025.9	84	-999	2.3	-1.1	-1.4	-999	0.3	6	0.1	0	eor
1980-12-29	403	5	4.8	7.4	8.0	1024.3	87	-999	5.7	1.4	0.4	-999	0.2	6	0.0	0	eor
1980-12-30	403	5	6.8	8.4	8.0	1021.5	85	-999	7.9	4.7	4.0	-999	0.0	6	0.0	0	eor
1980-12-31	403	5	6.2	7.0	7.7	1008.2	73	-999	7.4	5.3	5.3	-999	3.7	6	0.0	0	eor

4018 rows × 17 columns

Basic descriptive statistics¶

In [12]:

df1.describe()

Out[12]:

	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE
count	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	23741	23741.000000	23741.000000	23741.000000	23741	23741.000000	23741.000000	23741.000000	23741.000000
mean	7.189756	9.228912	9.489082	5.205307	1007.985151	76.475301	-999	13.351135	5.269264	3.511322	-999	1.623912	3.093678	-1.455907	0.817741
std	2.643762	7.641879	7.712391	2.255881	9.136360	14.642451	0	8.967447	6.763194	14.702692	0	3.851226	3.103872	78.322974	3.437469
min	3.000000	-17.900000	-999.000000	0.000000	961.000000	-999.000000	-999	-16.400000	-21.800000	-999.000000	-999	0.000000	0.000000	-999.000000	0.000000
25%	5.000000	3.400000	6.300000	3.700000	1002.400000	68.000000	-999	6.200000	0.500000	-0.900000	-999	0.000000	0.000000	0.300000	0.000000
50%	5.000000	9.400000	8.900000	5.700000	1008.300000	78.000000	-999	13.600000	5.400000	3.800000	-999	0.000000	1.000000	3.600000	0.000000
75%	10.000000	15.300000	12.500000	7.300000	1013.900000	86.000000	-999	20.400000	10.600000	9.000000	-999	1.500000	6.000000	7.900000	0.000000
max	10.000000	29.500000	23.900000	8.000000	1040.000000	100.000000	-999	37.800000	21.300000	21.000000	-999	106.000000	8.000000	16.367000	49.000000

access a single column

In [13]:

df1.SCHNEEHOEHE.describe().to_frame()

Out[13]:

	SCHNEEHOEHE
count	23741.000000
mean	0.817741
std	3.437469
min	0.000000
25%	0.000000
50%	0.000000
75%	0.000000
max	49.000000

Selecting Data¶

In [14]:

df1["SCHNEEHOEHE"] > 0

Out[14]:

MESS_DATUM
1950-01-01    False
1950-01-02    False
...
2014-12-30    True
2014-12-31    True
Name: SCHNEEHOEHE, Length: 23741

In [15]:

snow_days = df1[df1["SCHNEEHOEHE"] > 0]
snow_days

Out[15]:

	STATIONS_ID	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE	eor
MESS_DATUM
1950-01-06	403	5	2.6	5.6	7.3	997.50	79.00	-999	5.0	-4.0	-4.0	-999	7.2	8	-999.000	12	eor
1950-01-11	403	5	-7.2	3.1	7.0	1025.20	82.00	-999	-4.4	-8.7	-5.7	-999	0.6	1	-999.000	2	eor
1950-02-02	403	5	-5.0	3.1	8.0	1017.80	74.00	-999	-3.5	-6.9	-7.7	-999	0.0	0	-999.000	2	eor
1950-02-03	403	5	-2.0	3.5	7.3	1013.60	67.00	-999	1.2	-8.0	-9.4	-999	0.0	0	-999.000	2	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2014-02-06	403	3	4.5	6.4	4.7	995.82	77.25	-999	10.4	0.7	-2.5	-999	0.3	6	5.283	1	eor
2014-12-29	403	3	-3.6	4.5	5.6	1018.05	95.92	-999	-1.9	-6.0	-13.8	-999	0.0	7	6.117	11	eor
2014-12-30	403	3	-1.8	5.2	7.9	1020.31	97.42	-999	1.1	-3.5	-4.6	-999	0.6	8	0.000	8	eor
2014-12-31	403	3	2.8	7.4	8.0	1019.90	99.00	-999	4.9	1.1	0.0	-999	0.9	6	0.000	6	eor

2517 rows × 17 columns

In [16]:

snow_days.describe()

Out[16]:

	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE
count	2517.000000	2517.000000	2517.000000	2517.000000	2517.000000	2517.000000	2517	2517.000000	2517.000000	2517.000000	2517	2517.000000	2517.000000	2517.000000	2517.000000
mean	7.028208	-2.119031	4.593921	5.885936	1007.632932	83.989245	-999	0.376480	-4.994438	-6.662495	-999	1.211085	5.256257	-3.809342	7.713151
std	2.583502	4.070684	1.452455	2.300667	11.765287	9.092571	0	3.990027	4.586212	5.421881	0	2.347022	3.156976	77.125569	7.634561
min	3.000000	-17.900000	1.100000	0.000000	967.900000	44.000000	-999	-16.400000	-21.800000	-31.600000	-999	0.000000	0.000000	-999.000000	1.000000
25%	5.000000	-4.500000	3.500000	4.700000	999.300000	79.000000	-999	-1.900000	-7.900000	-9.800000	-999	0.000000	1.000000	0.000000	2.000000
50%	5.000000	-1.500000	4.500000	6.800000	1008.000000	85.290000	-999	0.700000	-4.200000	-5.600000	-999	0.100000	7.000000	0.400000	5.000000
75%	10.000000	0.600000	5.600000	7.800000	1016.300000	90.630000	-999	2.800000	-1.400000	-2.400000	-999	1.300000	7.000000	3.900000	11.000000
max	10.000000	12.700000	11.300000	8.000000	1040.000000	100.000000	-999	18.400000	8.800000	8.000000	-999	21.900000	8.000000	11.667000	49.000000

In [17]:

snow_days["2000-01-01":"2014-01-01"]

Out[17]:

	STATIONS_ID	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE	eor
MESS_DATUM
2000-01-21	403	10	-2.0	4.1	4.7	1006.90	73.00	-999	4.5	-3.8	-0.7	-999	0.3	7	4.70	4	eor
2000-01-22	403	10	-3.1	4.1	7.7	1000.50	83.00	-999	-1.6	-4.4	-6.9	-999	2.4	7	0.00	4	eor
2000-01-23	403	10	-6.1	2.6	3.0	1006.10	69.00	-999	-2.6	-8.7	-13.2	-999	0.1	7	7.60	6	eor
2000-01-24	403	10	-6.4	3.1	5.7	1014.40	77.00	-999	-1.9	-8.6	-9.1	-999	0.0	7	3.10	6	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2013-04-07	403	3	2.3	4.4	2.0	1009.70	64.67	-999	9.2	-4.6	-5.8	-999	0.0	6	8.75	6	eor
2013-04-08	403	3	3.7	4.6	4.9	1002.02	61.13	-999	9.6	-3.2	-4.8	-999	0.0	0	7.55	4	eor
2013-12-07	403	3	0.9	5.2	5.5	1008.05	80.46	-999	1.6	-0.1	-1.6	-999	0.1	8	0.40	1	eor
2013-12-08	403	3	3.6	7.5	8.0	1010.05	93.79	-999	7.3	0.0	-0.7	-999	5.3	8	0.00	2	eor

562 rows × 17 columns

Advanced Pandas

"Chengdu-pandas-d10". Licensed under CC BY-SA 2.5 es via Wikimedia Commons.

Combine 2 DataFrames¶

In [18]:

df1

Out[18]:

	STATIONS_ID	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE	eor
MESS_DATUM
1950-01-01	403	5	-3.2	4.0	5.0	1025.60	83.00	-999	-1.1	-4.9	-6.3	-999	2.2	7	-999.000	0	eor
1950-01-02	403	5	1.0	6.1	8.0	1005.60	95.00	-999	2.2	-3.7	-5.3	-999	12.6	8	-999.000	0	eor
1950-01-03	403	5	2.8	6.5	5.0	996.60	86.00	-999	3.9	1.7	-1.4	-999	0.5	1	-999.000	0	eor
1950-01-04	403	5	-0.1	5.2	7.7	999.50	85.00	-999	2.1	-0.9	-2.3	-999	0.5	7	-999.000	0	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2014-12-28	403	3	-5.1	3.8	4.6	1012.22	90.21	-999	-2.3	-7.6	-10.1	-999	3.6	7	1.333	0	eor
2014-12-29	403	3	-3.6	4.5	5.6	1018.05	95.92	-999	-1.9	-6.0	-13.8	-999	0.0	7	6.117	11	eor
2014-12-30	403	3	-1.8	5.2	7.9	1020.31	97.42	-999	1.1	-3.5	-4.6	-999	0.6	8	0.000	8	eor
2014-12-31	403	3	2.8	7.4	8.0	1019.90	99.00	-999	4.9	1.1	0.0	-999	0.9	6	0.000	6	eor

23741 rows × 17 columns

In [19]:

df2

Out[19]:

	STATIONS_ID	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE	eor
MESS_DATUM
1949-05-01	1443	5	8.5	8.1	8.0	986.40	76.00	-999.0	14.7	4.8	4.0	-999.0	0.0	0	0.0	0	eor
1949-05-02	1443	5	13.2	9.9	7.7	984.30	66.00	-999.0	15.2	9.5	9.0	-999.0	0.0	0	0.5	0	eor
1949-05-03	1443	5	16.7	13.2	3.0	980.40	72.00	-999.0	22.0	9.3	6.9	-999.0	0.0	0	12.6	0	eor
1949-05-04	1443	5	17.3	14.1	6.7	974.20	72.00	-999.0	22.6	10.6	9.0	-999.0	0.3	1	8.8	0	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2014-12-28	1443	3	-2.4	4.6	6.3	987.32	87.50	3.8	1.5	-9.6	-12.1	10.2	0.2	4	0.0	0	eor
2014-12-29	1443	3	-6.0	3.6	5.6	1005.85	89.30	1.4	-2.7	-11.7	-13.6	5.3	3.6	4	0.0	-999	eor
2014-12-30	1443	3	-1.7	5.1	7.9	1008.12	95.13	1.1	-0.1	-3.5	-3.6	4.4	1.2	4	0.0	5	eor
2014-12-31	1443	3	-0.7	5.6	7.8	1007.05	96.33	0.9	1.0	-4.8	-5.2	2.9	0.0	4	0.0	6	eor

23986 rows × 17 columns

The Index¶

In [20]:

df1.index

Out[20]:

<class 'pandas.tseries.index.DatetimeIndex'>
[1950-01-01, ..., 2014-12-31]
Length: 23741, Freq: None, Timezone: None

Be careful: some pandas functionality expects a sorted DatetimeIndex without enforcing it!

First Way¶

Concatenate

(see the excellent documentation at: http://pandas.pydata.org/pandas-docs/stable/merging.html)

In [21]:

df = pd.concat([df1, df2], keys=['Berlin', 'Freiburg'], axis=1)

In [22]:

df

Out[22]:

	Berlin										...	Freiburg
	STATIONS_ID	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	...	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE	eor
MESS_DATUM
1949-05-01	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	-999.0	14.7	4.8	4.0	-999.0	0.0	0	0.0	0	eor
1949-05-02	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	-999.0	15.2	9.5	9.0	-999.0	0.0	0	0.5	0	eor
1949-05-03	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	-999.0	22.0	9.3	6.9	-999.0	0.0	0	12.6	0	eor
1949-05-04	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	-999.0	22.6	10.6	9.0	-999.0	0.3	1	8.8	0	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2014-12-28	403	3	-5.1	3.8	4.6	1012.22	90.21	-999	-2.3	-7.6	...	3.8	1.5	-9.6	-12.1	10.2	0.2	4	0.0	0	eor
2014-12-29	403	3	-3.6	4.5	5.6	1018.05	95.92	-999	-1.9	-6.0	...	1.4	-2.7	-11.7	-13.6	5.3	3.6	4	0.0	-999	eor
2014-12-30	403	3	-1.8	5.2	7.9	1020.31	97.42	-999	1.1	-3.5	...	1.1	-0.1	-3.5	-3.6	4.4	1.2	4	0.0	5	eor
2014-12-31	403	3	2.8	7.4	8.0	1019.90	99.00	-999	4.9	1.1	...	0.9	1.0	-4.8	-5.2	2.9	0.0	4	0.0	6	eor

23986 rows × 34 columns

Now we got a hierarchichal index http://pandas.pydata.org/pandas-docs/stable/advanced.html#advanced

In [23]:

df = df.swaplevel(0,1, axis=1)

In [24]:

df

Out[24]:

	STATIONS_ID	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	...	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE	eor
	Berlin	Berlin	Berlin	Berlin	Berlin	Berlin	Berlin	Berlin	Berlin	Berlin	...	Freiburg	Freiburg	Freiburg	Freiburg	Freiburg	Freiburg	Freiburg	Freiburg	Freiburg	Freiburg
MESS_DATUM
1949-05-01	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	-999.0	14.7	4.8	4.0	-999.0	0.0	0	0.0	0	eor
1949-05-02	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	-999.0	15.2	9.5	9.0	-999.0	0.0	0	0.5	0	eor
1949-05-03	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	-999.0	22.0	9.3	6.9	-999.0	0.0	0	12.6	0	eor
1949-05-04	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	-999.0	22.6	10.6	9.0	-999.0	0.3	1	8.8	0	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2014-12-28	403	3	-5.1	3.8	4.6	1012.22	90.21	-999	-2.3	-7.6	...	3.8	1.5	-9.6	-12.1	10.2	0.2	4	0.0	0	eor
2014-12-29	403	3	-3.6	4.5	5.6	1018.05	95.92	-999	-1.9	-6.0	...	1.4	-2.7	-11.7	-13.6	5.3	3.6	4	0.0	-999	eor
2014-12-30	403	3	-1.8	5.2	7.9	1020.31	97.42	-999	1.1	-3.5	...	1.1	-0.1	-3.5	-3.6	4.4	1.2	4	0.0	5	eor
2014-12-31	403	3	2.8	7.4	8.0	1019.90	99.00	-999	4.9	1.1	...	0.9	1.0	-4.8	-5.2	2.9	0.0	4	0.0	6	eor

23986 rows × 34 columns

Alternative way¶

In [25]:

df = pd.concat([df1, df2], keys=['Berlin', 'Freiburg'])

In [26]:

df

Out[26]:

		STATIONS_ID	QUALITAETS_NIVEAU	LUFTTEMPERATUR	DAMPFDRUCK	BEDECKUNGSGRAD	LUFTDRUCK_STATIONSHOEHE	REL_FEUCHTE	WINDGESCHWINDIGKEIT	LUFTTEMPERATUR_MAXIMUM	LUFTTEMPERATUR_MINIMUM	LUFTTEMP_AM_ERDB_MINIMUM	WINDSPITZE_MAXIMUM	NIEDERSCHLAGSHOEHE	NIEDERSCHLAGSHOEHE_IND	SONNENSCHEINDAUER	SCHNEEHOEHE	eor
	MESS_DATUM
Berlin	1950-01-01	403	5	-3.2	4.0	5.0	1025.60	83.00	-999.0	-1.1	-4.9	-6.3	-999.0	2.2	7	-999	0	eor
	1950-01-02	403	5	1.0	6.1	8.0	1005.60	95.00	-999.0	2.2	-3.7	-5.3	-999.0	12.6	8	-999	0	eor
	1950-01-03	403	5	2.8	6.5	5.0	996.60	86.00	-999.0	3.9	1.7	-1.4	-999.0	0.5	1	-999	0	eor
	1950-01-04	403	5	-0.1	5.2	7.7	999.50	85.00	-999.0	2.1	-0.9	-2.3	-999.0	0.5	7	-999	0	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
Freiburg	2014-12-28	1443	3	-2.4	4.6	6.3	987.32	87.50	3.8	1.5	-9.6	-12.1	10.2	0.2	4	0	0	eor
	2014-12-29	1443	3	-6.0	3.6	5.6	1005.85	89.30	1.4	-2.7	-11.7	-13.6	5.3	3.6	4	0	-999	eor
	2014-12-30	1443	3	-1.7	5.1	7.9	1008.12	95.13	1.1	-0.1	-3.5	-3.6	4.4	1.2	4	0	5	eor
	2014-12-31	1443	3	-0.7	5.6	7.8	1007.05	96.33	0.9	1.0	-4.8	-5.2	2.9	0.0	4	0	6	eor

47727 rows × 17 columns

Fun with our MultiIndex¶

In [27]:

df = df.unstack(level=0)

In [28]:

df

Out[28]:

	STATIONS_ID		QUALITAETS_NIVEAU		LUFTTEMPERATUR		DAMPFDRUCK		BEDECKUNGSGRAD		...	NIEDERSCHLAGSHOEHE		NIEDERSCHLAGSHOEHE_IND		SONNENSCHEINDAUER		SCHNEEHOEHE		eor
	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	...	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg
MESS_DATUM
1949-05-01	NaN	1443	NaN	5	NaN	8.5	NaN	8.1	NaN	8.0	...	NaN	0.0	NaN	0	NaN	0.0	NaN	0	NaN	eor
1949-05-02	NaN	1443	NaN	5	NaN	13.2	NaN	9.9	NaN	7.7	...	NaN	0.0	NaN	0	NaN	0.5	NaN	0	NaN	eor
1949-05-03	NaN	1443	NaN	5	NaN	16.7	NaN	13.2	NaN	3.0	...	NaN	0.0	NaN	0	NaN	12.6	NaN	0	NaN	eor
1949-05-04	NaN	1443	NaN	5	NaN	17.3	NaN	14.1	NaN	6.7	...	NaN	0.3	NaN	1	NaN	8.8	NaN	0	NaN	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2014-12-28	403	1443	3	3	-5.1	-2.4	3.8	4.6	4.6	6.3	...	3.6	0.2	7	4	1.333	0.0	0	0	eor	eor
2014-12-29	403	1443	3	3	-3.6	-6.0	4.5	3.6	5.6	5.6	...	0.0	3.6	7	4	6.117	0.0	11	-999	eor	eor
2014-12-30	403	1443	3	3	-1.8	-1.7	5.2	5.1	7.9	7.9	...	0.6	1.2	8	4	0.000	0.0	8	5	eor	eor
2014-12-31	403	1443	3	3	2.8	-0.7	7.4	5.6	8.0	7.8	...	0.9	0.0	6	4	0.000	0.0	6	6	eor	eor

23986 rows × 34 columns

But data for Freiburg starts mid 49, we want only full years, so we just "say":

In [29]:

df = df.truncate(before="1950-01-01")

In [30]:

df

Out[30]:

	STATIONS_ID		QUALITAETS_NIVEAU		LUFTTEMPERATUR		DAMPFDRUCK		BEDECKUNGSGRAD		...	NIEDERSCHLAGSHOEHE		NIEDERSCHLAGSHOEHE_IND		SONNENSCHEINDAUER		SCHNEEHOEHE		eor
	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	...	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg
MESS_DATUM
1950-01-01	403	1443	5	5	-3.2	-2.9	4.0	4.3	5.0	1.3	...	2.2	0.0	7	0	-999.000	4.6	0	0	eor	eor
1950-01-02	403	1443	5	5	1.0	2.3	6.1	3.7	8.0	5.7	...	12.6	6.2	8	8	-999.000	1.0	0	0	eor	eor
1950-01-03	403	1443	5	5	2.8	2.4	6.5	6.5	5.0	8.0	...	0.5	6.2	1	8	-999.000	0.0	0	0	eor	eor
1950-01-04	403	1443	5	5	-0.1	2.4	5.2	6.8	7.7	8.0	...	0.5	9.0	7	8	-999.000	0.0	0	0	eor	eor
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2014-12-28	403	1443	3	3	-5.1	-2.4	3.8	4.6	4.6	6.3	...	3.6	0.2	7	4	1.333	0.0	0	0	eor	eor
2014-12-29	403	1443	3	3	-3.6	-6.0	4.5	3.6	5.6	5.6	...	0.0	3.6	7	4	6.117	0.0	11	-999	eor	eor
2014-12-30	403	1443	3	3	-1.8	-1.7	5.2	5.1	7.9	7.9	...	0.6	1.2	8	4	0.000	0.0	8	5	eor	eor
2014-12-31	403	1443	3	3	2.8	-0.7	7.4	5.6	8.0	7.8	...	0.9	0.0	6	4	0.000	0.0	6	6	eor	eor

23741 rows × 34 columns

Descriptive again¶

In [31]:

df.describe()

Out[31]:

	QUALITAETS_NIVEAU		LUFTTEMPERATUR		DAMPFDRUCK		BEDECKUNGSGRAD		LUFTDRUCK_STATIONSHOEHE		...	WINDSPITZE_MAXIMUM		NIEDERSCHLAGSHOEHE		NIEDERSCHLAGSHOEHE_IND		SONNENSCHEINDAUER		SCHNEEHOEHE
	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	...	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg
count	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	...	23741	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000	23741.000000
mean	7.189756	6.287014	9.228912	10.897077	9.489082	9.409667	5.205307	-109.661084	1007.985151	982.044536	...	-999	-68.607527	1.623912	2.378960	3.093678	2.327071	-1.455907	4.585664	0.817741	-127.160777
std	2.643762	2.726823	7.641879	11.992317	7.712391	25.693188	2.255881	319.922454	9.136360	45.276124	...	0	270.671541	3.851226	14.065206	3.103872	13.290091	78.322974	16.561068	3.437469	333.418686
min	3.000000	3.000000	-17.900000	-999.000000	-999.000000	-999.000000	0.000000	-999.000000	961.000000	-999.000000	...	-999	-999.000000	0.000000	-999.000000	0.000000	-999.000000	-999.000000	-999.000000	0.000000	-999.000000
25%	5.000000	5.000000	3.400000	5.300000	6.300000	6.700000	3.700000	2.400000	1002.400000	978.200000	...	-999	6.000000	0.000000	0.000000	0.000000	0.000000	0.300000	0.400000	0.000000	0.000000
50%	5.000000	5.000000	9.400000	11.300000	8.900000	9.400000	5.700000	5.700000	1008.300000	983.400000	...	-999	8.900000	0.000000	0.000000	1.000000	1.000000	3.600000	3.900000	0.000000	0.000000
75%	10.000000	10.000000	15.300000	16.900000	12.500000	13.100000	7.300000	7.300000	1013.900000	988.300000	...	-999	12.400000	1.500000	2.700000	6.000000	6.000000	7.900000	8.400000	0.000000	0.000000
max	10.000000	10.000000	29.500000	30.800000	23.900000	23.100000	8.000000	8.000000	1040.000000	1012.700000	...	-999	36.400000	106.000000	80.300000	8.000000	8.000000	16.367000	15.200000	49.000000	40.000000

8 rows × 30 columns

But the -999 value is clearly not valid, we replace it with "NaN"

In [32]:

df.replace(-999, np.nan, inplace=True)

In [33]:

df.describe()

Out[33]:

	QUALITAETS_NIVEAU		LUFTTEMPERATUR		DAMPFDRUCK		BEDECKUNGSGRAD		LUFTDRUCK_STATIONSHOEHE		...	WINDSPITZE_MAXIMUM		NIEDERSCHLAGSHOEHE		NIEDERSCHLAGSHOEHE_IND		SONNENSCHEINDAUER		SCHNEEHOEHE
	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	...	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg	Berlin	Freiburg
count	23741.000000	23741.000000	23741.000000	23739.000000	23740.000000	23726.000000	23741.000000	21021.000000	23741.000000	23729.000000	...	0	21889.000000	23741.000000	23737.000000	23741.000000	23737.000000	23596.000000	23735.000000	23741.000000	20712.000000
mean	7.189756	6.287014	9.228912	10.982160	9.531563	10.047201	5.205307	5.414405	1007.985151	983.046371	...	NaN	10.111778	1.623912	2.547706	3.093678	2.495808	4.674110	4.839362	0.817741	0.340238
std	2.643762	2.726823	7.641879	7.608645	4.079052	4.150623	2.255881	2.365689	9.136360	8.074512	...	NaN	4.674865	3.851226	5.371165	3.103872	2.767641	4.402997	4.433495	3.437469	1.839556
min	3.000000	3.000000	-17.900000	-15.400000	1.100000	1.200000	0.000000	0.000000	961.000000	939.200000	...	NaN	0.300000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000
25%	5.000000	5.000000	3.400000	5.300000	6.300000	6.700000	3.700000	3.700000	1002.400000	978.200000	...	NaN	6.700000	0.000000	0.000000	0.000000	0.000000	0.400000	0.400000	0.000000	0.000000
50%	5.000000	5.000000	9.400000	11.300000	8.900000	9.400000	5.700000	6.000000	1008.300000	983.400000	...	NaN	9.300000	0.000000	0.000000	1.000000	1.000000	3.700000	3.900000	0.000000	0.000000
75%	10.000000	10.000000	15.300000	16.900000	12.500000	13.100000	7.300000	7.300000	1013.900000	988.300000	...	NaN	12.800000	1.500000	2.700000	6.000000	6.000000	7.900000	8.400000	0.000000	0.000000
max	10.000000	10.000000	29.500000	30.800000	23.900000	23.100000	8.000000	8.000000	1040.000000	1012.700000	...	NaN	36.400000	106.000000	80.300000	8.000000	8.000000	16.367000	15.200000	49.000000	40.000000

8 rows × 30 columns

From now on¶

From now on we consider only "Lufttemperatur"

In [34]:

airtemp = df.LUFTTEMPERATUR

In [35]:

airtemp.describe()

Out[35]:

	Berlin	Freiburg
count	23741.000000	23739.000000
mean	9.228912	10.982160
std	7.641879	7.608645
min	-17.900000	-15.400000
25%	3.400000	5.300000
50%	9.400000	11.300000
75%	15.300000	16.900000
max	29.500000	30.800000

Plotting¶

In [36]:

airtemp.plot(fontsize=16)

Out[36]:

<matplotlib.axes._subplots.AxesSubplot at 0x7effcd806bd0>

In [37]:

ax = airtemp.boxplot(return_type='axes', fontsize=20)

Year's average - Groupby¶

In [38]:

dfyear=airtemp.groupby(airtemp.index.year).mean()
dfyear

Out[38]:

	Berlin	Freiburg
1950	9.330685	10.961370
1951	9.680274	10.835342
1952	8.423770	10.627322
1953	10.038082	10.667671
...	...	...
2011	10.043288	11.530137
2012	9.545902	10.943443
2013	9.395890	10.352603
2014	11.044384	11.874795

65 rows × 2 columns

In [39]:

dfyear.plot()

Out[39]:

<matplotlib.axes._subplots.AxesSubplot at 0x7effccd07ad0>

In [40]:

pd.rolling_mean(dfyear, 10).plot()

Out[40]:

<matplotlib.axes._subplots.AxesSubplot at 0x7effccf1ed50>

Number of days with snow¶

In [41]:

snow_days.groupby(snow_days.index.year)['SCHNEEHOEHE'].count().plot()

Out[41]:

<matplotlib.axes._subplots.AxesSubplot at 0x7effccf13090>

In [42]:

snow_days.groupby(snow_days.index.year)['SCHNEEHOEHE'].count().to_frame()[-7:]

Out[42]:

	SCHNEEHOEHE
2008	14
2009	52
2010	93
2011	19
2012	36
2013	69
2014	19

Statistics

Correlations¶

In [43]:

airtemp.corr()

Out[43]:

	Berlin	Freiburg
Berlin	1.00000	0.91505
Freiburg	0.91505	1.00000

Statsmodels¶

In [44]:

import statsmodels.formula.api as smf
import statsmodels.api as sm

In [45]:

dfyear['year'] = dfyear.index
dfyear

Out[45]:

	Berlin	Freiburg	year
1950	9.330685	10.961370	1950
1951	9.680274	10.835342	1951
1952	8.423770	10.627322	1952
1953	10.038082	10.667671	1953
...	...	...	...
2011	10.043288	11.530137	2011
2012	9.545902	10.943443	2012
2013	9.395890	10.352603	2013
2014	11.044384	11.874795	2014

65 rows × 3 columns

In [46]:

berlin_model = smf.ols(formula="Berlin ~ year", data=dfyear).fit()
freiburg_model = smf.ols(formula="Freiburg ~ year", data=dfyear).fit()

In [47]:

berlin_model.summary()

Out[47]:

OLS Regression Results
Dep. Variable:	Berlin	R-squared:	0.233
Model:	OLS	Adj. R-squared:	0.221
Method:	Least Squares	F-statistic:	19.18
Date:	Thu, 13 Aug 2015	Prob (F-statistic):	4.59e-05
Time:	15:28:39	Log-Likelihood:	-70.925
No. Observations:	65	AIC:	145.8
Df Residuals:	63	BIC:	150.2
Df Model:	1
Covariance Type:	nonrobust

	coef	std err	t	P>\|t\|	[95.0% Conf. Int.]
Intercept	-32.7638	9.590	-3.416	0.001	-51.928 -13.600
year	0.0212	0.005	4.379	0.000	0.012 0.031

Omnibus:	1.993	Durbin-Watson:	1.610
Prob(Omnibus):	0.369	Jarque-Bera (JB):	1.916
Skew:	-0.342	Prob(JB):	0.384
Kurtosis:	2.510	Cond. No.	2.09e+05

In [48]:

freiburg_model.summary()

Out[48]:

OLS Regression Results
Dep. Variable:	Freiburg	R-squared:	0.278
Model:	OLS	Adj. R-squared:	0.267
Method:	Least Squares	F-statistic:	24.29
Date:	Thu, 13 Aug 2015	Prob (F-statistic):	6.32e-06
Time:	15:28:39	Log-Likelihood:	-68.552
No. Observations:	65	AIC:	141.1
Df Residuals:	63	BIC:	145.5
Df Model:	1
Covariance Type:	nonrobust

	coef	std err	t	P>\|t\|	[95.0% Conf. Int.]
Intercept	-34.5893	9.246	-3.741	0.000	-53.067 -16.112
year	0.0230	0.005	4.929	0.000	0.014 0.032

Omnibus:	1.153	Durbin-Watson:	1.289
Prob(Omnibus):	0.562	Jarque-Bera (JB):	1.149
Skew:	-0.301	Prob(JB):	0.563
Kurtosis:	2.753	Cond. No.	2.09e+05

In [49]:

import seaborn as sns
sns.set(style="ticks")

# Load the example dataset for Anscombe's quartet
df = sns.load_dataset("anscombe")

# Show the results of a linear regression within each dataset
sns.lmplot(x="year", y="Berlin", data=dfyear,
           size=8,
           scatter_kws={"s": 50, "alpha": 1})
plt.plot(dfyear.index, dfyear.Berlin)

Out[49]:

[<matplotlib.lines.Line2D at 0x7effc08bdbd0>]

In [50]:

sns.lmplot(x="year", y="Freiburg", data=dfyear,
           size=8,
           scatter_kws={"s": 50, "alpha": 1})
plt.plot(dfyear.index, dfyear.Freiburg)

Out[50]:

[<matplotlib.lines.Line2D at 0x7effbec9ee10>]

Thanks for your attention!

Have fun!

"Xiao Liwu im San Diego Zoo - Foto 3" by User:jballeis - Own work. Licensed under CC BY-SA 3.0 via Wikimedia Commons.

Questions?

Bonus

In [51]:

df.columns

Out[51]:

Index([u'dataset', u'x', u'y'], dtype='object')

In [52]:

gg=airtemp.groupby([airtemp.index.month, airtemp.index.year]).mean()

In [53]:

gg

Out[53]:

		Berlin	Freiburg
1	1950	-1.383871	0.948387
	1951	1.096774	4.222581
	1952	1.187097	1.590323
	1953	0.667742	-1.019355
...	...	...	...
12	2011	4.370968	5.990323
	2012	0.554839	4.183871
	2013	4.248387	3.154839
	2014	2.751613	4.841935

780 rows × 2 columns

In [54]:

gg.loc[1].plot()

Out[54]:

<matplotlib.axes._subplots.AxesSubplot at 0x7effbe4d1690>

In [55]:

gg.loc[10].plot()

Out[55]:

<matplotlib.axes._subplots.AxesSubplot at 0x7effbcc21290>

In [56]:

gg=airtemp.groupby([airtemp.index.month]).mean()
gg.plot()

Out[56]:

<matplotlib.axes._subplots.AxesSubplot at 0x7effcd289350>

In [57]:

gg=airtemp.groupby([airtemp.index.dayofyear]).mean()
gg.plot()

Out[57]:

<matplotlib.axes._subplots.AxesSubplot at 0x7effbcaa4dd0>