pipeline updated

malleswarigelli · Apr 8, 2024 · e3dc432 · e3dc432
1 parent 5dcc19a
commit e3dc432
Show file tree

Hide file tree

Showing 8 changed files with 64 additions and 54 deletions.
diff --git a/.dockerignore b/.dockerignore
@@ -8,4 +8,5 @@ demo.py
 README.md
 LICENSE
 us_visa.egg.info
-notebook
+notebook
+credentials.txt
diff --git a/.github/workflows/aws.yaml b/.github/workflows/aws.yaml
@@ -54,4 +54,4 @@ jobs:
 
       - name: Run Docker Image to serve users
         run: |
-         docker run -d -e AWS_ACCESS_KEY_ID="${{ secrets.AWS_ACCESS_KEY_ID }}" -e AWS_SECRET_ACCESS_KEY="${{ secrets.AWS_SECRET_ACCESS_KEY }}" -e AWS_DEFAULT_REGION="${{ secrets.AWS_DEFAULT_REGION }}" -e MONGODB_URL="${{ secrets.MONGODB_URL }}" -p 8080:8080 "${{ steps.login-ecr.outputs.registry }}"/"${{ secrets.ECR_REPO }}":latest
+         docker run -d -e AWS_ACCESS_KEY_ID="${{ secrets.AWS_ACCESS_KEY_ID }}" -e AWS_SECRET_ACCESS_KEY="${{ secrets.AWS_SECRET_ACCESS_KEY }}" -e AWS_DEFAULT_REGION="${{ secrets.AWS_DEFAULT_REGION }}" -e MONGODB_URL="${{ secrets.MONGODB_URL }}" -p 1111:1111 "${{ steps.login-ecr.outputs.registry }}"/"${{ secrets.ECR_REPO }}":latest
diff --git a/.gitignore b/.gitignore
@@ -159,4 +159,5 @@ cython_debug/
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
 
-artifact/*
+artifact/*
+credentials.txt
diff --git a/README.md b/README.md
@@ -57,7 +57,7 @@ os.getenv(MONGODB_URL_KEY)
 
 ```run this command in gitbash to set MONGODB_URL
 
-export MONGODB_URL= "mongodb+srv://username:password@cluster0.ppivwau.mongodb.net/?retryWrites=true&w=majority&appName=Cluster0"
+export MONGODB_URL="mongodb+srv://username:password@cluster0.ppivwau.mongodb.net/?retryWrites=true&w=majority&appName=Cluster0"
 export AWS_ACCESS_KEY_ID=<AWS_ACCESS_KEY_ID>
 export AWS_SECRET_ACCESS_KEY=<AWS_SECRET_ACCESS_KEY>
 
@@ -93,7 +93,7 @@ export AWS_SECRET_ACCESS_KEY=<AWS_SECRET_ACCESS_KEY>
 2. AmazonEC2FullAccess
 
 ## 3. Create ECR repo to store/save docker image
-- Save the URI: 136566696263.dkr.ecr.us-east-1.amazonaws.com/mlproject
+- Save the URI: 637423357032.dkr.ecr.us-east-1.amazonaws.com/usvisa_ecr
 
 ## 4. Create EC2 machine (Ubuntu)
 

diff --git a/US_Visa/constant/__init__.py b/US_Visa/constant/__init__.py
@@ -64,9 +64,9 @@
 MODEL EVALUATION related constant start with MODEL_EVALUATION var name
 """
 MODEL_EVALUATION_CHANGED_THRESHOLD_SCORE: float = 0.02
-MODEL_BUCKET_NAME = "usvisamodel2024"
+MODEL_BUCKET_NAME = "usvisamodel24"
 MODEL_PUSHER_S3_KEY = "model-registry"
 
 
 APP_HOST = "0.0.0.0"
-APP_PORT = 8080
+APP_PORT = 1111
diff --git a/US_Visa/pipeline/prediction_pipeline.py b/US_Visa/pipeline/prediction_pipeline.py
@@ -42,20 +42,7 @@ def __init__(self,
 
 
         except Exception as e:
-            raise USvisaException(e, sys) from e
-
-    def get_usvisa_input_data_frame(self)-> DataFrame:
-        """
-        This function returns a DataFrame from USvisaData class input
-        """
-        try:
-
-            usvisa_input_dict = self.get_usvisa_data_as_dict()
-            return DataFrame(usvisa_input_dict)
-
-        except Exception as e:
-            raise USvisaException(e, sys) from e
-
+            raise USvisaException(e, sys) from e  
 
     def get_usvisa_data_as_dict(self):
         """
@@ -85,6 +72,18 @@ def get_usvisa_data_as_dict(self):
 
         except Exception as e:
             raise USvisaException(e, sys) from e
+
+    def get_usvisa_input_data_frame(self)-> DataFrame:
+        """
+        This function converts dictionary from get_usvisa_data_as_dict into a pandas dataframe
+        """
+        try:
+
+            usvisa_input_dict = self.get_usvisa_data_as_dict()
+            return DataFrame(usvisa_input_dict)
+
+        except Exception as e:
+            raise USvisaException(e, sys) from e
 
 class USvisaClassifier:
     def __init__(self,prediction_pipeline_config: USvisaPredictorConfig = USvisaPredictorConfig(),) -> None:
@@ -97,7 +96,7 @@ def __init__(self,prediction_pipeline_config: USvisaPredictorConfig = USvisaPred
         except Exception as e:
             raise USvisaException(e, sys)
 
-    def predict(self, dataframe) -> str:
+    def predict(self, df:DataFrame) -> str:
         """
         This is the method of USvisaClassifier
         Returns: Prediction in string format
@@ -108,7 +107,7 @@ def predict(self, dataframe) -> str:
                 bucket_name=self.prediction_pipeline_config.model_bucket_name,
                 model_path=self.prediction_pipeline_config.model_file_path,
             )
-            result =  model.predict(dataframe)
+            result =  model.predict(df)
 
             return result
 

diff --git a/app.py b/app.py
@@ -99,7 +99,7 @@ async def predictRouteClient(request: Request):
 
         model_predictor = USvisaClassifier()
 
-        value = model_predictor.predict(dataframe=usvisa_df)[0]
+        value = model_predictor.predict(df=usvisa_df)[0]
 
         status = None
         if value == 1:

diff --git a/notebook/mongoDB_demo.ipynb b/notebook/mongoDB_demo.ipynb
@@ -32,7 +32,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 19,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -41,7 +41,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 20,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [
     {
@@ -182,7 +182,7 @@
        "4      149907.3900         Year                  Y   Certified  "
       ]
      },
-     "execution_count": 20,
+     "execution_count": 2,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -194,7 +194,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 21,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [
     {
@@ -203,7 +203,7 @@
        "(25480, 12)"
       ]
      },
-     "execution_count": 21,
+     "execution_count": 3,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -214,7 +214,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 22,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
@@ -227,7 +227,7 @@
        "      dtype='object')"
       ]
      },
-     "execution_count": 22,
+     "execution_count": 4,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -238,7 +238,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 23,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [
     {
@@ -12247,7 +12247,7 @@
        " ...]"
       ]
      },
-     "execution_count": 23,
+     "execution_count": 5,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -12259,7 +12259,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -12270,7 +12270,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 14,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -12279,7 +12279,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 15,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -12291,16 +12291,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 16,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "<pymongo.cursor.Cursor at 0x16679617c40>"
+       "<pymongo.cursor.Cursor at 0x20c2006aac0>"
       ]
      },
-     "execution_count": 4,
+     "execution_count": 16,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -12312,7 +12312,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 17,
    "metadata": {},
    "outputs": [
     {
@@ -37809,7 +37809,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 18,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -37819,7 +37819,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 19,
    "metadata": {},
    "outputs": [
     {
@@ -37973,7 +37973,7 @@
        "4   Certified  "
       ]
      },
-     "execution_count": 8,
+     "execution_count": 19,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -37984,19 +37984,11 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 20,
    "metadata": {},
    "outputs": [],
    "source": [
-    "## If you are getting timeout issue\n",
-    "\n",
-    "import certifi\n",
-    "client = pymongo.MongoClient(CONNECTION_URL, tlsCAFile=certifi.where())\n",
-    "# TO CREATE THE DATABASE\n",
-    "data_base = client[DB_NAME]\n",
-    "collection = data_base[COLLECTION_NAME]\n",
-    "# TO INSERT DATA INTO THE COLLECTION\n",
-    "#rec = collection.insert_many(data)\n"
+    "\n"
    ]
   },
   {
@@ -38241,14 +38233,31 @@
     "print(from_root)"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "## If you are getting timeout issue\n",
+    "\n",
+    "import certifi\n",
+    "client = pymongo.MongoClient(CONNECTION_URL, tlsCAFile=certifi.where())\n",
+    "# TO CREATE THE DATABASE\n",
+    "data_base = client[DB_NAME]\n",
+    "collection = data_base[COLLECTION_NAME]\n",
+    "# TO INSERT DATA INTO THE COLLECTION\n",
+    "rec = collection.insert_many(data)"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
     "# run this command in gitbash to set MONGODB_URL while running app.py or demo.py file \n",
-    "# export MONGODB_URL= \"mongodb+srv://username:password@cluster0.ppivwau.mongodb.net/?retryWrites=true&w=majority&appName=Cluster0\""
+    "# export MONGODB_URL=\"mongodb+srv://username:password@cluster0.ppivwau.mongodb.net/?retryWrites=true&w=majority&appName=Cluster0\""
    ]
   }
  ],