Xnew, Pnew = kf.filter_update(filtered_state_mean=Xold,
filtered_state_covariance=Pold,
observation=ys[t])
Pold = np.array(Pnew)
Xold = np.array(Xnew)
y[t - burn_in] = np.dot(Xnew, H)
# ---------------------------------------- #
# In[170]:
# ---------- Plot ------------------------ #
plt.subplot(211)
plt.plot(ts, ys, label='noisy') # noisy measures
plt.plot(ts[burn_in:Ns], y, label='filtered') # filtered measures
plt.grid()
plt.legend(bbox_to_anchor=(0., 1.02, 1., .102),
loc=3,
ncol=2,
mode="expand",
borderaxespad=0.)
plt.subplot(212)
residuals = ys.ravel()[burn_in:Ns] - y.ravel()
plt.plot(ts[burn_in:Ns], residuals)
plt.title('Durbin-Watson: {:.3f}'.format(sm.stats.durbin_watson(residuals)))
plt.grid()
plt.tight_layout()
plt.show()
# ---------------------------------------- #
# In[ ]:
df = pd.concat(frames)
df['ncalls'] = df['ncalls'].apply(lambda x: x.split('/')[0])
df['ncalls'] = df['ncalls'].astype(int)
df = df[df[function_colname].str.contains("constph|openmm")]
df = df.loc[df['Attempts/Update'] == 1]
# df = df.loc[df[function_colname] == "constph.py:1035(update)"]
# Initialize a grid of plots with an Axes for each walk
# sns.swarmplot(data=df, x='NCMC steps', y='percall', hue=function_colname, ax=ax)
for fname, fgroup in df.groupby(function_colname):
f, ax = plt.subplots(figsize=(7, 7))
ax.set_title(fname)
ax.scatter(fgroup["NCMC steps"], fgroup["percall.1"])
# sns.factorplot(data=df, x='NCMC steps', y='percall.1', hue=function_colname,ax=ax)
plt.show()