def fit_serveral_genes(series, fitter, loo, filename, b_show):
if fitter is not None:
theta, L, LOO_predictions,_ = fitter.fit(series.ages, series.expression, loo=loo)
print 'L = {}'.format(L)
fig = plot_series(series, fitter.shape, theta, LOO_predictions)
else:
fig = plot_series(series)
if filename is None:
ensure_dir(results_dir())
filename = join(results_dir(), 'fits.png')
print 'Saving figure to {}'.format(filename)
save_figure(fig, filename)
if b_show:
plt.show(block=True)
def fit_serveral_genes(series, fitter, loo, filename, b_show):
if fitter is not None:
theta, L, LOO_predictions, _ = fitter.fit(series.ages,
series.expression,
loo=loo)
print 'L = {}'.format(L)
fig = plot_series(series, fitter.shape, theta, LOO_predictions)
else:
fig = plot_series(series)
if filename is None:
ensure_dir(results_dir())
filename = join(results_dir(), 'fits.png')
print 'Saving figure to {}'.format(filename)
save_figure(fig, filename)
if b_show:
plt.show(block=True)
def transform(self, X):
return self._gauss_basis(X[:, :, np.newaxis],
self._centres,
self._width,
axis=1)
def linear(train, test, t=132):
X_train, X_test = sklearn_formatting(train, test)
gauss_model = make_pipeline(
GaussianFeatures(40),
LinearRegression(),
)
gauss_model.fit(X_train, train.values)
y_fit = gauss_model.predict(X_train)
# predict a cycle
y_pred = gauss_model.predict(X_test)
rmse = error(test.values, y_pred)
return y_fit, y_pred, rmse
if __name__ == "__main__":
df_train, df_test = get_data()
lin_y, lin_y_pred, lin_rmse = linear(df_train, df_test)
plot_series(df_train, df_test, lin_y, lin_y_pred)
def arima_fn(train, test, order=(10, 1, 9)):
# fit ARIMA with calculated order from above
model = ARIMA(train.values, order=order)
res = model.fit(disp=-1)
y_fit = pd.Series(
res.fittedvalues,
copy=True,
index=train.index[1:],
).cumsum()
# predict a cycle
y_pred = res.predict(
start=len(train),
end=len(train) + len(test) - 1,
).cumsum()
rmse = error(test.values, y_pred)
return y_fit, y_pred, rmse
if __name__ == "__main__":
df_train, df_test = get_data()
df_train = stationary(df_train)
df_test = stationary(df_test)
arima_y, arima_y_pred, arima_rmse = arima_fn(df_train, df_test)
plot_series(df_train.iloc[1:], df_test, arima_y, arima_y_pred)
for i, g in enumerate(series.gene_names):
print 'Fitting series {}...'.format(i + 1)
theta, sigma, LOO_predictions, _ = fitter.fit(x, y[:, i], loo=True)
fit = Bunch(
theta=theta,
LOO_predictions=LOO_predictions,
)
fits.append(fit)
print 'Fitting with correlations...'
levels = fitter.fit_multi(x, y, loo=True, n_iterations=2)
res = levels[-1]
print 'Theta:'
for ti in res.theta:
print ' {}'.format(ti)
print 'Sigma:'
print res.sigma
plot_series(series, fitter.shape, res.theta, res.LOO_predictions)
R2_pairs = []
for i, g in enumerate(series.gene_names):
y_real = y[:, i]
y_basic = fits[i].LOO_predictions
y_multi_gene = res.LOO_predictions[:,
i] # no NANs in the generated data, so no need to handle the original_inds mess
basic_R2 = loo_score(y_real, y_basic)
multi_gene_R2 = loo_score(y_real, y_multi_gene)
R2_pairs.append((basic_R2, multi_gene_R2))
plot_comparison_scatter(R2_pairs, series.region_name)
print 'R2_pairs = {}'.format(R2_pairs)
fits = []
for i,g in enumerate(series.gene_names):
print 'Fitting series {}...'.format(i+1)
theta, sigma, LOO_predictions,_ = fitter.fit(x,y[:,i],loo=True)
fit = Bunch(
theta = theta,
LOO_predictions = LOO_predictions,
)
fits.append(fit)
print 'Fitting with correlations...'
levels = fitter.fit_multi(x, y, loo=True, n_iterations=2)
res = levels[-1]
print 'Theta:'
for ti in res.theta:
print ' {}'.format(ti)
print 'Sigma:'
print res.sigma
plot_series(series, fitter.shape, res.theta, res.LOO_predictions)
R2_pairs = []
for i,g in enumerate(series.gene_names):
y_real = y[:,i]
y_basic = fits[i].LOO_predictions
y_multi_gene = res.LOO_predictions[:,i] # no NANs in the generated data, so no need to handle the original_inds mess
basic_R2 = loo_score(y_real,y_basic)
multi_gene_R2 = loo_score(y_real,y_multi_gene)
R2_pairs.append( (basic_R2, multi_gene_R2) )
plot_comparison_scatter(R2_pairs,series.region_name)
print 'R2_pairs = {}'.format(R2_pairs)
from sklearn.gaussian_process.kernels import RBF, WhiteKernel, ExpSineSquared
from plots import plot_series
from sunspots import error, get_data, sklearn_formatting
def gp(train, test, t=132):
X_train, X_test = sklearn_formatting(train, test)
gp_kernel = 2**2 \
+ ExpSineSquared(1, 60000.0) \
+ ExpSineSquared(2, 120000.0) \
+ WhiteKernel(2.5)
gpr = GaussianProcessRegressor(kernel=gp_kernel)
gpr.fit(X_train, train.values)
y_fit = gpr.predict(X_train, return_std=False)
# predict a cycle
y_pred = gpr.predict(X_test, return_std=False)
rmse = error(test.values, y_pred)
return y_fit, y_pred, rmse
if __name__ == "__main__":
df_train, df_test = get_data()
gauss_y, gauss_y_pred, gauss_rmse = gp(df_train, df_test)
plot_series(df_train, df_test, gauss_y, gauss_y_pred)