# Define filtros sobre los campos del avaluo (query con la estructura (campo1=value11 or campo1=value12 or ... ) AND (campo2=value21 or campo2=value22 or ... ) filters = {"codigo_postal_ubicacion_inmueble": ['03103', '03100', '03104', '03200', '03230', '03240']} #filters = {"codigo_postal_ubicacion_inmueble": ['03103', '03100', '03104', '03200', '03230', '03240', '03023', '03000','03020', '03023', '03600']} #filters = {"codigo_postal_ubicacion_inmueble": ['77710', '77712', '77713', '77714', '77716', '77717', '77718', '77720', '77723', '77724', '77725', '77726' ]} #filters = {"codigo_postal_ubicacion_inmueble": ['77710']} # Avaluos como una lista de diccionarios appraisals = data_manager.get_inputs_dicionaries(file_name+'.csv', filters) num_appraisals = len(appraisals) generated_ints = random.sample(xrange(0,num_appraisals), num_appraisals) appraisals_random_sorted = [appraisals[i] for i in generated_ints] # creacion del mapa de avaluos filtrados visualizer.dot_infomap_from_appraisals(appraisals, "map_filtered_"+file_name) # salia a consula del primer avaluo formated_list = visualizer.list_formated_appraisal(appraisals[0]) print "----------------------" print "-Ejemplo de avaluo----" for each_field in formated_list: print "Campo: ", each_field print "----------------------" # selecciona el 80-20 para entrenamiento y pruebas
# filters = {"codigo_postal_ubicacion_inmueble": ['77710', '77712', '77713', '77714', '77716', '77717', '77718', '77720', '77723', '77724', '77725', '77726' ]} # filters = {"codigo_postal_ubicacion_inmueble": ['77710']} # filters = {"codigo_postal_ubicacion_inmueble": ['14410']} filter_string_msg = "-" + "-".join(filters["codigo_postal_ubicacion_inmueble"]) # Avaluos como una lista de diccionarios appraisals = data_manager.get_inputs_dicionaries(file_name_examples + ".csv", filters) num_appraisals = len(appraisals) generated_ints = random.sample(xrange(0, num_appraisals), num_appraisals) appraisals_random_sorted = [appraisals[i] for i in generated_ints] # creacion del mapa de avaluos filtrados visualizer.dot_infomap_from_appraisals(appraisals, "map_filtered_" + file_name_examples + filter_string_msg) # salia a consula del primer avaluo formated_list = visualizer.list_formated_appraisal(appraisals[0]) print "----------------------" print "-Ejemplo de avaluo----" for each_field in formated_list: print "Campo: ", each_field print "----------------------" # Resultados por correlacion cruzada 10-fold-cross-correlation hidden_size = 50 k = 10 data_slice = num_appraisals // k
file_name = "quintanaroo_abril_2015" data_dict_avaluos = load_csv(file_name+'.csv') for i in data_dict_avaluos[:2]: print i.keys() print visualizer.dot_infomap_from_appraisals(data_dict_avaluos, "UVs_map_"+file_name)