Merge pull request #80 from fboundy/70-syntax-error-reported-in-pv_optpy

load from daily

apps/pv_opt/config/config.yaml

@@ -23,6 +23,7 @@ pv_opt:
   # inverter_type: "SOLIS_CORE_MODBUS"
   # inverter_type: SOLIS_SOLARMAN
   device_name: solis
+  consumption_history_days: 2
 
   # ========================================
   # Plant parameters
@@ -125,7 +126,7 @@ pv_opt:
   # battery_voltage: sensor.{device_name}_battery_voltage
   # maximum_dod_percent: sensor.{device_name}_overdischarge_soc
   # update_cycle_seconds: 60
-
+  # id_consumption_today: sensor.{device_name}_house_load_today
   # id_consumption:
   #   - sensor.{device_name}_house_load_power
   #   - sensor.{device_name}_backup_load_power

apps/pv_opt/pv_opt.py

@@ -19,7 +19,7 @@
 #
 USE_TARIFF = True
 
-VERSION = "3.4.5-alpha"
+VERSION = "3.4.5-beta-02"
 DEBUG = True
 
 DATE_TIME_FORMAT_LONG = "%Y-%m-%d %H:%M:%S%z"
@@ -231,17 +231,17 @@ def importName(modulename, name):
 class PVOpt(hass.Hass):
     def hass2df(self, entity_id, days=2, log=False):
         if log:
-            self.log(f">>> Getting {days} history for {entity_id}")
+            self.log(f">>> Getting {days} days' history for {entity_id}")
             self.log(f">>> Entity exits: {self.entity_exists(entity_id)}")
         hist = self.get_history(entity_id=entity_id, days=days)
 
-        if log:
-            self.log(f">>> {hist}")
+        # if log:
+        #     self.log(f">>> {hist}")
         df = pd.DataFrame(hist[0]).set_index("last_updated")["state"]
         df.index = pd.to_datetime(df.index, format="ISO8601")
 
-        if log:
-            self.log(f">>> {df}")
+        # if log:
+        #     self.log(f">>> {df}")
 
         df = df.sort_index()
         df = df[df != "unavailable"]
@@ -1762,30 +1762,47 @@ def load_consumption(self):
         self.log("Getting expected consumption data")
 
         consumption = pd.Series(index=self.static.index, data=0)
-        for entity_id in self.config["id_consumption"]:
-            try:
-                df = self.hass2df(
-                    entity_id,
-                    days=int(self.get_config("consumption_history_days")),
-                    log=self.debug,
-                )
 
-            except Exception as e:
-                self.log(
-                    f"Unable to get historical consumption from {entity_id}. {e}",
-                    level="ERROR",
-                )
-                return False
+        if self.entity_exists(self.config["id_consumption_today"]):
+            entity_ids = self.config["id_consumption_today"]
+            consumption_daily = True
+        else:
+            entity_ids = self.config["id_consumption"]
+            consumption_daily = False
 
-            df.index = pd.to_datetime(df.index)
-            df = (
-                pd.to_numeric(df, errors="coerce")
-                .dropna()
-                .resample("30T")
-                .mean()
-                .fillna(0)
+        if not isinstance(entity_ids, list):
+            entity_ids = [entity_ids]
+
+        for entity_id in entity_ids:
+            df = self.hass2df(
+                entity_id,
+                days=int(self.get_config("consumption_history_days")),
+                log=self.debug,
             )
 
+            # except Exception as e:
+            #     self.log(
+            #         f"Unable to get historical consumption from {entity_id}. {e}",
+            #         level="ERROR",
+            #     )
+            #     return False
+
+            df.index = pd.to_datetime(df.index)
+
+            if consumption_daily:  #
+                df = pd.to_numeric(df, errors="coerce")
+                df = (
+                    df.diff(-1).fillna(0).clip(upper=0).cumsum().resample("30T")
+                ).ffill().fillna(0).diff(-1) * 2000
+            else:
+                df = (
+                    pd.to_numeric(df, errors="coerce")
+                    .dropna()
+                    .resample("30T")
+                    .mean()
+                    .fillna(0)
+                )
+
             df = df * (1 + self.get_config("consumption_margin") / 100)
             dfx = pd.Series(index=df.index, data=df.to_list())
             # Group by time and take the mean

apps/pv_opt/pvpy.py

@@ -511,6 +511,8 @@ def flows(self, initial_soc, static_flows, slots=[], soc_now=None, **kwargs):
 
         df["chg"] = chg[:-1]
         df["chg"] = df["chg"].ffill()
+        df["chg_end"] = chg[1:]
+        df["chg_end"] = df["chg_end"].bfill()
         df["battery"] = (pd.Series(chg).diff(-1) / freq)[:-1].to_list()
         df.loc[df["battery"] > 0, "battery"] = (
             df["battery"] * self.inverter.inverter_efficiency
@@ -521,7 +523,7 @@ def flows(self, initial_soc, static_flows, slots=[], soc_now=None, **kwargs):
         df["grid"] = -(solar - consumption + df["battery"]).round(0)
         df["forced"] = forced_charge
         df["soc"] = (df["chg"] / self.battery.capacity) * 100
-        df["soc_end"] = df["soc"].shift(-1)
+        df["soc_end"] = (df["chg_end"] / self.battery.capacity) * 100
 
         return df
 

apps/pv_opt/solis.py

@@ -32,6 +32,7 @@
         "default_config": {
             "maximum_dod_percent": "number.{device_name}_battery_minimum_soc",
             "id_battery_soc": "sensor.{device_name}_battery_soc",
+            "id_consumption_today": "sensor.{device_name}_house_load_today",
             "id_consumption": [
                 "sensor.{device_name}_house_load",
                 "sensor.{device_name}_bypass_load",