Updating regression model & import_model fixes

Author: smlindauer

examples/pzmmModelImportExampleRegression.ipynb renamed to examples/ARCHIVE/v1_8/pzmmModelImportExampleRegression.ipynb

@@ -1,16 +1,17 @@
 {
     "name": "LinearRegression",
-    "description": "",
-    "function": "classification",
+    "description": "Description for the LinearRegression model.",
     "scoreCodeType": "python",
-    "trainTable": " ",
+    "trainTable": "",
     "trainCodeType": "Python",
     "algorithm": "",
+    "function": "Prediction",
     "targetVariable": "Price",
-    "targetEvent": null,
-    "targetLevel": "BINARY",
-    "eventProbVar": "EM_PREDICTION",
+    "targetEvent": "",
+    "targetLevel": "INTERVAL",
+    "eventProbVar": "",
     "modeler": "sasdemo",
     "tool": "Python 3",
-    "toolVersion": "3.8.3"
+    "toolVersion": "3.8.16",
+    "properties": []
 }

examples/data/USAHousingModels/LinearRegression/LinearRegression.pickle

@@ -9,7 +9,7 @@
     },
     {
         "role": "score",
-        "name": "LinearRegressionScore.py"
+        "name": "score_LinearRegression.py"
     },
     {
         "role": "scoreResource",

examples/data/USAHousingModels/LinearRegression/LinearRegression.zip

@@ -1,32 +1,32 @@
 [
     {
         "name": "Avg_Area_Income",
-        "length": 8,
+        "level": "interval",
         "type": "decimal",
-        "level": "interval"
+        "length": 8
     },
     {
         "name": "Avg_Area_House_Age",
-        "length": 8,
+        "level": "interval",
         "type": "decimal",
-        "level": "interval"
+        "length": 8
     },
     {
         "name": "Avg_Area_Number_of_Rooms",
-        "length": 8,
+        "level": "interval",
         "type": "decimal",
-        "level": "interval"
+        "length": 8
     },
     {
         "name": "Avg_Area_Number_of_Bedrooms",
-        "length": 8,
+        "level": "interval",
         "type": "decimal",
-        "level": "interval"
+        "length": 8
     },
     {
         "name": "Area_Population",
-        "length": 8,
+        "level": "interval",
         "type": "decimal",
-        "level": "interval"
+        "length": 8
     }
 ]

examples/data/USAHousingModels/LinearRegression/ModelProperties.json

@@ -1,8 +1,8 @@
 [
     {
         "name": "EM_PREDICTION",
-        "length": 8,
+        "level": "interval",
         "type": "decimal",
-        "level": "interval"
+        "length": 8
     }
 ]

examples/data/USAHousingModels/LinearRegression/fileMetadata.json

@@ -0,0 +1,58 @@
+import math
+import pickle
+import pandas as pd
+import numpy as np
+from pathlib import Path
+
+import settings
+
+with open(Path(settings.pickle_path) / "LinearRegression.pickle", "rb") as pickle_model:
+    model = pickle.load(pickle_model)
+
+def score(Avg_Area_Income, Avg_Area_House_Age, Avg_Area_Number_of_Rooms, Avg_Area_Number_of_Bedrooms, Area_Population):
+    "Output: EM_PREDICTION"
+
+    try:
+        global model
+    except NameError:
+        with open(Path(settings.pickle_path) / "LinearRegression.pickle", "rb") as pickle_model:
+                model = pickle.load(pickle_model)
+
+    try:
+        if math.isnan(Avg_Area_Income):
+            Avg_Area_Income = 68583.10898397
+    except TypeError:
+        Avg_Area_Income = 68583.10898397
+    try:
+        if math.isnan(Avg_Area_House_Age):
+            Avg_Area_House_Age = 5.977222035287
+    except TypeError:
+        Avg_Area_House_Age = 5.977222035287
+    try:
+        if math.isnan(Avg_Area_Number_of_Rooms):
+            Avg_Area_Number_of_Rooms = 6.9877918509092005
+    except TypeError:
+        Avg_Area_Number_of_Rooms = 6.9877918509092005
+    try:
+        if math.isnan(Avg_Area_Number_of_Bedrooms):
+            Avg_Area_Number_of_Bedrooms = 3.9813300000000003
+    except TypeError:
+        Avg_Area_Number_of_Bedrooms = 3.9813300000000003
+    try:
+        if math.isnan(Area_Population):
+            Area_Population = 36163.516038540256
+    except TypeError:
+        Area_Population = 36163.516038540256
+
+    input_array = pd.DataFrame([[Avg_Area_Income, Avg_Area_House_Age, Avg_Area_Number_of_Rooms, Avg_Area_Number_of_Bedrooms, Area_Population]],
+                              columns=["Avg_Area_Income", "Avg_Area_House_Age", "Avg_Area_Number_of_Rooms", "Avg_Area_Number_of_Bedrooms", "Area_Population"],
+                              dtype=float)
+    prediction = model.predict(input_array)
+
+    # Check for numpy values and convert to a CAS readable representation
+    if isinstance(prediction, np.ndarray):
+        prediction = prediction.tolist()[0]
+
+    EM_PREDICTION = prediction
+
+    return EM_PREDICTION