def write_comparison_csv(file_name: str,
data,
model: Model,
parameters: Union[list, dict],
model_dt: float,
headers: Optional[List[str]] = None):
if isinstance(parameters, dict):
parameters = [parameters]
x0, y0 = find_t0_data_point(data)
t_max = find_max_t_data_point(data)[0]
x_data = [[t, xt] for t, xt, yt in data]
y_data = [[t, yt] for t, xt, yt in data]
x_datas = []
y_datas = []
for parameters0 in parameters:
model_data = model.get_data(parameters0, x0, y0, 0, t_max,
math.ceil(t_max / model_dt + 1))
x_datas.append([[t, xt] for t, xt, yt in model_data])
y_datas.append([[t, yt] for t, xt, yt in model_data])
final_data = x_data
for model_x_data in x_datas:
final_data = unite_data(final_data, model_x_data)
final_data = unite_data(final_data, y_data)
for model_y_data in y_datas:
final_data = unite_data(final_data, model_y_data)
new_headers = None if headers is None else [""] + [
"x " + h for h in headers
] + ["y " + h for h in headers]
export_array_array_to_csv(final_data, file_name, headers=new_headers)
def s_squared(data, model: Model, parameters, x_factor=1, y_factor=1):
# find datapoint for t=0
x0, y0 = find_t0_data_point(data)
x, y = model.get_prediction(parameters, x0, y0)
return sum((x(t) - xt)**2 * x_factor + (y(t) - yt)**2 * y_factor
for t, xt, yt in data) / len(data)
def write_comparison_csv(file_name: str, data, model: Model, parameters: Union[list, dict], model_dt: float):
if isinstance(parameters, dict):
parameters = [parameters]
t0, x0, y0 = find_t0_data_point(data)
t_max = find_max_t_data_point(data)[0]
for parameters0 in parameters:
model_data = model.get_data(parameters0, x0, y0, 0, t_max, math.ceil(t_max/model_dt+1))
data = unite_data(data, model_data)
export_data_to_svg(data, file_name)
def s_squared(data, model: Model, parameters, e1, e2):
# find datapoint for t=0
x0, y0 = find_t0_data_point(data)
x, y = model.get_prediction(parameters, x0, y0)
try:
return sum((math.log(max(x(t), 0)+e1)-math.log(max(xt, 0)+e1))**2 +
(math.log(max(y(t), 0)+e2)-math.log(max(yt, 0)+e2))**2 for t, xt, yt in data)/len(data)
except ValueError:
raise ModelException("a value below e1 or e2 occurred!")
y_factor=y_factor, error_out=single_data_errors[2*i+1],
debug_mod=debug_mod,
minimal_second_gradient=minimal_second_gradient)
# write to the output file content string
result_file_content += "Using data from {}:\n".format(data_file_names[i])
result_file_content += "Normal gradient descent found the following parameters that gave error {}:\n {}\n".format(
single_data_errors[2*i][-1], str(new_parameters1)
)
result_file_content += "Improved gradient descent found the following parameters that gave error {}:\n {}\n".format(
single_data_errors[2*i + 1][-1], str(new_parameters2)
)
result_file_content += "\n"
# write some files that are used to create graphs to visually compare the model
x0, y0 = find_t0_data_point(datas[i])
t_max = find_max_t_data_point(datas[i])[0]
model_data1 = model.get_data(new_parameters1, x0, y0, 0, t_max, math.ceil(t_max/model_graph_dt+1))
model_data2 = model.get_data(new_parameters2, x0, y0, 0, t_max, math.ceil(t_max/model_graph_dt+1))
model_prey_data1 = [data_point[:2] for data_point in model_data1]
model_prey_data2 = [data_point[:2] for data_point in model_data2]
model_predator_data1 = [[data_point[0], data_point[2]] for data_point in model_data1]
model_predator_data2 = [[data_point[0], data_point[2]] for data_point in model_data2]
# write prey graph
prey_graph_data = [data_point[:2] for data_point in datas[i]]
prey_graph_data = unite_data(prey_graph_data, model_prey_data1)
prey_graph_data = unite_data(prey_graph_data, model_prey_data2)
export_array_array_to_csv(prey_graph_data, output_folder + "/" + single_data_set_prey_graph_format.format(i+1))
# write predator graph