import pandas as pd import numpy as np import matplotlib.pyplot as plt pd.set_option("display.max_rows", 8) plt.rcParams['figure.figsize'] = (9, 6)

dates = pd.date_range('20130101', periods=6) pd.DataFrame(np.random.randn(6,4), index=dates, columns=list('ABCD'))

pd.DataFrame({'A' : 1., 'B' : pd.Timestamp('20130102'), 'C' : pd.Series(1,index=list(range(4)),dtype='float32'), 'D' : np.arange(4,dtype='int32'), 'E' : pd.Categorical(["test","train","test","train"]), 'F' : 'foo' })

url = "https://www.fun-mooc.fr/c4x/agrocampusouest/40001S03/asset/AnaDo_JeuDonnees_TemperatFrance.csv" french_cities = pd.read_csv(url, delimiter=";", encoding="latin1", index_col=0) french_cities

french_cities.head()

french_cities.tail()

french_cities.index

french_cities.index.name = "City" french_cities.head()

import locale import calendar locale.setlocale(locale.LC_ALL,'C') months = calendar.month_abbr print(*months) french_cities.rename( columns={ old : new for old, new in zip(french_cities.columns[:12], months[1:]) if old != new }, inplace=True) french_cities.rename(columns={'Moye':'Mean'}, inplace=True) french_cities

# Needs `lxml`, `beautifulSoup4` and `html5lib` python packages table_list = pd.read_html("http://www.psmsl.org/data/obtaining/") table_list

# there is 1 table on that page which contains metadata about the stations where # sea levels are recorded local_sea_level_stations = table_list[0] local_sea_level_stations

french_cities['Lati'] # DF [] accesses columns (Series)

french_cities.Lati

french_cities.loc[:,'Lati']

french_cities.loc['Rennes', "Dec"]

french_cities.loc['Rennes', ["Sep", "Dec"]]

french_cities.loc['Rennes', "Sep":"Dec"]

mask = [True, False] * 6 + 5 * [False] mask

mask = [True, False] * 6 + 5 * [False] print(french_cities.iloc[:, mask])

print(french_cities.loc["Rennes", mask])

french_cities["std"] = french_cities.iloc[:,:12].std(axis=1) french_cities

french_cities = french_cities.drop("std", axis=1) # remove this new column

french_cities

# french_cities['Rennes']['Sep'] = 25 # It does not works and breaks the DataFrame french_cities.loc['Rennes']['Sep'] # = 25 is the right way to do it

french_cities

french_cities['Mean'].min(), french_cities['Ampl'].max()

def fahrenheit(T): return T*9/5+32

list(map(lambda T: T*9/5+32, np.arange(20)))

fahrenheit = lambda T: T*9/5+32 fahrenheit(-15)

fahrenheit = lambda T: T*9/5+32 french_cities['Mean'].apply(fahrenheit)

french_cities.sort_values(by='Lati')

french_cities = french_cities.sort_values(by='Lati',ascending=False) french_cities

pd.set_option("display.max_rows", 20) unstacked = french_cities.iloc[:,:12].unstack() unstacked

type(unstacked)

city_temp = french_cities.iloc[:,:12].transpose() city_temp.plot()

city_temp.boxplot(rot=90);

french_cities['Région'].describe()

french_cities['Région'].unique()

french_cities['Région'].value_counts()

# To save memory, we can convert it to a categorical column: french_cities["Région"] = french_cities["Région"].astype("category")

french_cities.memory_usage()

fc_grouped_region = french_cities.groupby("Région") type(fc_grouped_region)

for group_name, subdf in fc_grouped_region: print(group_name) print(subdf) print("")

df = pd.read_excel("Folds5x2_pp.xlsx") df.corr()

import seaborn as sns sns.pairplot(df)