def test_cpp_extensions_preds():
statements = [(
("large", -1, "UN/entities/human/financial/economic/inflation"),
("small", 1, "UN/events/human/human_migration"),
)]
G = AnalysisGraph.from_causal_fragments(statements)
G.map_concepts_to_indicators()
G["UN/events/human/human_migration"].replace_indicator(
"Net migration", "New asylum seeking applicants", "UNHCR")
G.to_png()
# Now we can specify how to initialize betas. Posible values are:
# InitialBeta.ZERO
# InitialBeta.ONE
# InitialBeta.HALF
# InitialBeta.MEAN
# InitialBeta.RANDOM - A random value between [-1, 1]
G.train_model(2015,
1,
2015,
12,
10,
10,
initial_beta=InitialBeta.ZERO,
use_continuous=False)
preds = G.generate_prediction(2015, 1, 2016, 12)
pred_plot(preds, "New asylum seeking applicants", save_as="pred_plot.pdf")
def test_pred_compare(G_unit):
EN.train_model(G_unit, 2015, 1, 2015, 12, 1000, 1000, k=1)
EN.generate_predictions(G_unit, 2016, 1, 2016, 12)
pred_df = EN.mean_pred_to_df(G_unit, "Net migration", true_vals=True)
print(pred_df)
EN.pred_plot(G_unit, "Net migration", plot_type="Comparison")
def test_pred(G_unit):
EN.train_model(G_unit, 2015, 1, 2015, 12, 1000, 1000, k=1)
EN.generate_predictions(G_unit, 2016, 1, 2016, 12)
pred = EN.pred_to_array(G_unit, "Net migration")
print(pred)
pred_df = EN.mean_pred_to_df(G_unit, "Net migration")
print(pred_df)
EN.pred_plot(G_unit, "Net migration")
G.parameterize("South Sudan", "Jonglei", "", 2017, 4, {})
ind = G[
"wm/concept/indicator_and_reported_property/agriculture/Crop_Production"].get_indicator(
"Average Harvested Weight at Maturity (Maize)")
draw_CAG(G)
G.train_model(
2014,
6,
2016,
3,
country="South Sudan",
res=20,
burn=100,
use_heuristic=False,
)
G.set_default_initial_state()
G.s0[1] = 0.1
preds = G.generate_prediction(2016, 3, 2016, 7)
# preds = G.generate_prediction(2018, 1, 2018, 2)
EN.pred_plot(
preds,
"IPC Phase Classification",
0.95,
plot_type="Prediction",
show_rmse=True,
show_training_data=True,
use_heuristic_for_true=False,
save_as="Oct2019EvalPred.png",
)
def test_pred_error(G_unit):
EN.train_model(G_unit, 2015, 1, 2015, 12, 1000, 1000, k=1)
EN.generate_predictions(G_unit, 2016, 1, 2016, 12)
EN.pred_plot(G_unit, "Net migration", plot_type="Error")