#take I(2) series, etc.).
#not implemented since is not entirely needed for Granger test.
#2. difference once to I(1) series which are roughly stationary
#(e ~iid N(0,sigma^2))
#not implemented since is not entirely necesary for Granger test.
#3. calculate partial autocorrelation function (PACF) w/ Ljung-Box test to see
#which lags help predict current values.
clicksPACF = stattools.pacf_ols(clicksPerDay, nlags=MAX_LAG)
encountersPACF = stattools.pacf_ols(encountersPerDay, nlags=MAX_LAG)
#Ljung-Box test
clicksPACF_LJ = _math.ljungBox(clicksPACF, len(clicksPerDay), MAX_LAG)
encountersPACF_LJ = _math.ljungBox(encountersPACF, len(encountersPerDay), MAX_LAG)
print("clicks PACF Ljung-Box")
print(tabulate(clicksPACF_LJ, headers=["lag", "R", "Q", "p-val"]))
print("encounters PACF Ljung-Box")
print(tabulate(encountersPACF_LJ, headers=["lag", "R", "Q", "p-val"]))
#plot
def plotxy(title, x, y, LSRL=None):
fig = plt.figure()
plt.bar(x, y, width=0.08)
plt.grid()
plt.title(title)
fig.autofmt_xdate()
plt.show()
#clicksPerDay.index.name = None #encountersPerDay.index.name = None clicksPerDay = pd.Series(data=clicksPerDay["count_clicks"], index=clicksPerDay.index) encountersPerDay = pd.Series(data=encountersPerDay["count_encounter"], index=encountersPerDay.index) clicksPerDay = clicksPerDay.fillna(method="ffill") encountersPerDay = encountersPerDay.fillna(method="ffill") ####################LJUNG-BOX#################### clicksPACF = stattools.pacf_ols(clicksPerDay, nlags=MAX_LAG) encountersPACF = stattools.pacf_ols(encountersPerDay, nlags=MAX_LAG) #my implementation results = _math.ljungBox(encountersPACF, len(encountersPerDay), MAX_LAG) lag, R, Q, p = zip(*results) print np.asarray(Q[1:]) print np.asarray(p[1:]) #statsmodels implementation results = diagnostic.acorr_ljungbox(encountersPerDay, lags=MAX_LAG) print results #my copy of the statsmodels implementation results = _math.ljungBox2(encountersPerDay, maxlag=MAX_LAG) print results #they are not the same... I wonder why.
tval = tNum/tDen
df = len(R)-3
pval = scipy.stats.t.sf(tval, df)
header = ["type", "value"]
data = [
["tval",tval],
["pval",pval],
["df",df]
]
print("\n")
print("*William's Modified Hotelling T Test*")
print(tabulate.tabulate(data, headers=header))
#Ljung Box Test & Autocorrelation/Crosscorrelation
coeffs = _math.ljungBox(R, R, maxLag=7)
header = ["lag", "R value", "Q", "p value"]
data = coeffs
print("\n")
print("*Ljung-Box Test for Crosscorrelation*")
print("## R vs R ##")
print(tabulate.tabulate(data, headers=header))
coeffs = _math.ljungBox(t, t, maxLag=7)
header = ["lag", "R value", "Q", "p value"]
data = coeffs
print("## t vs t ##")
print(tabulate.tabulate(data, headers=header))
coeffs = _math.ljungBox(rsent, rsent, maxLag=7)
header = ["lag", "R value", "Q", "p value"]
