Spaces:
Sleeping
Sleeping
File size: 5,636 Bytes
3b66598 fff2149 3b66598 fff2149 3b66598 fff2149 3b66598 fff2149 3b66598 fff2149 3b66598 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 |
import json
import joblib
import pandas as pd
from sklearn.preprocessing import StandardScaler
from pickle import load
import numpy as np
class RTUPipeline:
scaler = None
def __init__(self, scaler_path=None):
self.output_col_names = [
"hp_hws_temp",
"rtu_003_sa_temp",
"rtu_003_oadmpr_pct",
"rtu_003_ra_temp",
"rtu_003_oa_temp",
"rtu_003_ma_temp",
"rtu_003_sf_vfd_spd_fbk_tn",
"rtu_003_rf_vfd_spd_fbk_tn",
"rtu_004_sa_temp",
"rtu_004_oadmpr_pct",
"rtu_004_ra_temp",
"rtu_004_oa_temp",
"rtu_004_ma_temp",
"rtu_004_sf_vfd_spd_fbk_tn",
"rtu_004_rf_vfd_spd_fbk_tn",
"rtu_001_sa_temp",
"rtu_001_oadmpr_pct",
"rtu_001_ra_temp",
"rtu_001_oa_temp",
"rtu_001_ma_temp",
"rtu_001_sf_vfd_spd_fbk_tn",
"rtu_001_rf_vfd_spd_fbk_tn",
"rtu_002_sa_temp",
"rtu_002_oadmpr_pct",
"rtu_002_ra_temp",
"rtu_002_oa_temp",
"rtu_002_ma_temp",
"rtu_002_sf_vfd_spd_fbk_tn",
"rtu_002_rf_vfd_spd_fbk_tn",
# "rtu_004_sat_sp_tn",
# "rtu_003_sat_sp_tn",
# "rtu_001_sat_sp_tn",
# "rtu_002_sat_sp_tn",
# "air_temp_set_1",
# "air_temp_set_2",
# "dew_point_temperature_set_1d",
# "relative_humidity_set_1",
# "solar_radiation_set_1",
]
self.input_col_names = ["air_temp_set_1",
"air_temp_set_2",
"dew_point_temperature_set_1d",
"relative_humidity_set_1",
"solar_radiation_set_1",
]
self.num_inputs = len(self.input_col_names)
self.num_outputs = len(self.output_col_names)
self.column_names = self.output_col_names+self.input_col_names
if scaler_path:
self.scaler = self.get_scaler(scaler_path)
self.df = pd.DataFrame(columns = self.column_names)
def get_scaler(self, scaler_path):
return scaler_path
def get_window(self, df):
len_df = np.len(df)
if len_df > 30:
return df[len_df - 31 : len_df].astype("float32")
else:
return None
def transform_window(self, df_window):
return self.scaler.transform(df_window)
def prepare_input(self,df_trans):
return df_trans[:30, :].reshape((1, 30, len(self.column_names)))
def extract_data_from_message(self, message):
payload = json.loads(message.payload.decode())
len_df = len(self.df)
# self.df.loc[len_df] = {'hp_hws_temp':payload['hp_hws_temp'],
# 'rtu_003_sa_temp':payload['rtu_003_sa_temp'],
# 'rtu_003_oadmpr_pct': payload["rtu_003_oadmpr_pct"],
# 'rtu_003_ra_temp':payload["rtu_003_ra_temp"],
# 'rtu_003_oa_temp': payload["rtu_003_oa_temp"],
# 'rtu_003_ma_temp': payload["rtu_003_ma_temp"],
# 'rtu_003_sf_vfd_spd_fbk_tn': payload["rtu_003_sf_vfd_spd_fbk_tn"],
# 'rtu_003_rf_vfd_spd_fbk_tn':payload["rtu_003_rf_vfd_spd_fbk_tn"],
# 'rtu_004_sa_temp':payload["rtu_004_sa_temp"],
# 'rtu_004_oadmpr_pct':payload["rtu_004_oadmpr_pct"],
# 'rtu_004_ra_temp':payload["rtu_004_ra_temp"],
# 'rtu_004_oa_temp':payload["rtu_004_oa_temp"],
# 'rtu_004_ma_temp':payload["rtu_004_ma_temp"],
# 'rtu_004_sf_vfd_spd_fbk_tn':payload["rtu_004_sf_vfd_spd_fbk_tn"],
# 'rtu_004_rf_vfd_spd_fbk_tn':payload["rtu_004_rf_vfd_spd_fbk_tn"],
# 'rtu_001_sa_temp':payload["rtu_001_sa_temp"],
# 'rtu_001_oadmpr_pct': payload["rtu_001_oadmpr_pct"],
# 'rtu_001_ra_temp':payload["rtu_001_ra_temp"],
# 'rtu_001_oa_temp': payload["rtu_001_oa_temp"],
# 'rtu_001_ma_temp': payload["rtu_001_ma_temp"],
# 'rtu_001_sf_vfd_spd_fbk_tn': payload["rtu_001_sf_vfd_spd_fbk_tn"],
# 'rtu_001_rf_vfd_spd_fbk_tn':payload["rtu_001_rf_vfd_spd_fbk_tn"],
# 'rtu_002_sa_temp':payload["rtu_002_sa_temp"],
# 'rtu_002_oadmpr_pct':payload["rtu_002_oadmpr_pct"],
# 'rtu_002_ra_temp':payload["rtu_002_ra_temp"],
# 'rtu_002_oa_temp':payload["rtu_002_oa_temp"],
# 'rtu_002_ma_temp':payload["rtu_002_ma_temp"],
# 'rtu_002_sf_vfd_spd_fbk_tn':payload["rtu_002_sf_vfd_spd_fbk_tn"],
# 'rtu_002_rf_vfd_spd_fbk_tn':payload["rtu_002_rf_vfd_spd_fbk_tn"],
# 'rtu_004_sat_sp_tn':payload["rtu_004_sat_sp_tn"],
# 'rtu_003_sat_sp_tn' :payload["rtu_003_sat_sp_tn"],
# 'rtu_001_sat_sp_tn':payload["rtu_001_sat_sp_tn"],
# 'rtu_002_sat_sp_tn':payload["rtu_002_sat_sp_tn"],
# 'air_temp_set_1':payload["air_temp_set_1"],
# 'air_temp_set_2':payload["air_temp_set_2"],
# 'dew_point_temperature_set_1d':payload["dew_point_temperature_set_1d"],
# 'relative_humidity_set_1':payload["relative_humidity_set_1"],
# 'solar_radiation_set_1':payload["solar_radiation_set_1"]}
self.df.loc[len_df] = {}
for col in self.column_names:
self.df.loc[len_df][col] = payload[col]
return self.df
def fit(self,message):
len_df = np.len(df)
df = self.extract_data_from_message(message)
df_window = self.get_window(df, len_df)
if df_window is not None:
df_trans = self.transform_window(df_window, self.scaler)
df_new = self.prepare_input(df_trans)
return df_new,df_trans
|