def calculate_scores(data):
if len(data) == 0:
return math.nan, math.nan
else:
#relative stock performance
data_pct = (data - data.iloc[0])/data.iloc[0]*100
#yearly stock performance
data_year = data.groupby([data.index.year]).apply(lambda x: (x.iloc[-1]-x.iloc[0])/x.iloc[0]*100)
year_median = data_year.median()
cagr = (((data_pct.iloc[-1] / 100 + 1) ** (1/data_year.index.size) - 1) * 100)
return cagr, year_median
#%%
# aggregating data with rolling mean
daily = data.resample('D').sum()
daily.rolling(30, center = True).sum().plot(style = [':', '--', '-'])
plt.ylabel('mean hourly count')
#%%
# Ussing Gaussian window
daily.rolling(50, center = True,
win_type = 'gaussian').sum(std = 10).plot()
#%%
# Digging in
# average traffic as a function of tge time of day
by_time = data.groupby(data.index.time).mean()
hourly_ticks = 4 * 60 * 60 * np.arange(6)
by_time.plot(xticks = hourly_ticks)
#%%
# How things change based on the day of the week
by_weekday = data.groupby(data.index.dayofweek).mean()
by_weekday.index = ['Mon', 'Tues', 'Wed', 'Thurs', 'Fri', 'Sat', 'Sun']
by_weekday.plot()
#%%
# let's do a compound groupby and look at the hourly trend on
# weekdats versus weekends
weekend = np.where(data.index.weekday < 5, 'Weekday', 'Weekend')
by_time = data.groupby([weekend, data.index.time]).mean()
data.to_csv("/Users/xiangliu/Desktop/CSC560 Data/pollution_AQI.csv",
index=True,
sep=',')
# In[6]:
data = pd.read_csv('/Users/xiangliu/Desktop/CSC560 Data/pollution_AQI.csv')
data.shape
# In[7]:
data.head(3)
# In[8]:
data.groupby(['State']).count()
#Found there are 5 states no there(Montana, Nebraska, Mississippi, West virgina, Vermont)
# In[9]:
le = data['CO AQI']
le[le.isnull()]
# In[10]:
le = data['SO2 AQI']
le[le.isnull()]
# In[11]:
le = data['NO2 AQI']