From SW's laptop

master
sanginnwoo 2022-10-14 20:07:26 +09:00
parent 03957cc7bc
commit 27998d1195
2 changed files with 14 additions and 8 deletions

View File

@ -21,11 +21,9 @@ for (root, directories, files) in os.walk(input_dir):
for input_file in input_files:
for i in range(20, 100, 20):
ERROR = settle_prediction_steps.run_settle_prediction(input_file,
ERROR = settle_prediction_steps_main.run_settle_prediction(input_file,
output_dir, i, 100, False, False)
df.loc[len(df.index)] = [input_file, i, ERROR[0], ERROR[1], ERROR[2],
ERROR[3], ERROR[4], ERROR[5]]
df.to_csv('Error.csv')

View File

@ -54,14 +54,20 @@ def run_settle_prediction(input_file, output_dir,
# 파일 읽기, 데이터 설정
# ====================
# 현재 파일 이름 출력
print("Working on " + input_file)
# CSV 파일 읽기
data = pd.read_csv(input_file)
data = pd.read_csv(input_file, encoding='euc-kr')
# 시간, 침하량, 성토고 배열 생성
time = data['Time'].to_numpy()
settle = data['Settle'].to_numpy()
settle = data['Settlement'].to_numpy()
surcharge = data['Surcharge'].to_numpy()
# 데이터 닫기
# 마지막 계측 데이터 index + 1 파악
final_index = time.size
@ -120,7 +126,7 @@ def run_settle_prediction(input_file, output_dir,
step_span = step_end_date - step_start_date
step_data_num = step_end_index[i] - step_start_index[i] + 1
if (step_span > 30 and step_data_num > 15):
if (step_span > 50 and step_data_num > 15):
step_start_index_adjust.append((step_start_index[i]))
step_end_index_adjust.append((step_end_index[i]))
@ -472,7 +478,7 @@ def run_settle_prediction(input_file, output_dir,
plt.title(filename + ": up to %i%% data used in the final step" % final_step_predict_percent)
# 그래프 저장 (SVG 및 PNG)
plt.savefig(output_dir + '/' + filename +' %i percent (SVG).svg' %final_step_predict_percent, bbox_inches='tight')
#plt.savefig(output_dir + '/' + filename +' %i percent (SVG).svg' %final_step_predict_percent, bbox_inches='tight')
plt.savefig(output_dir + '/' + filename +' %i percent (PNG).png' %final_step_predict_percent, bbox_inches='tight')
# 그래프 출력
@ -489,4 +495,6 @@ def run_settle_prediction(input_file, output_dir,
# 산정 에러값 반환
return [RMSE_hyper_original, RMSE_hyper_nonlinear, RMSE_step,
final_error_hyper_original, final_error_hyper_nonlinear,
final_error_step]
final_error_step]
run_settle_prediction('data/1_S-8.csv', 'output', 80, 100, False, False)