Merge pull request #29 from DushanSenadheera/dev

update api

model/data/southern.csv

@@ -1,11 +1,11 @@
-Title,Reviews,Rating,Category,Location,Duration,Budget
-Galle Fort,3516,4.5,point of interests,Galle,3,0
-Bentota Beach,2305,4.5,beaches,Bentota,1,0
-Old Town of Galle and its Fortifications,3511,4.5,point of interests,Galle,1,0
-Mirissa Beach,3095,4.5,beaches,Mirissa,1,0
-Sea Turtle Farm Galle Mahamodara,1076,4.5,nature & wildlife,Galle,2,30
-Hikkaduwa Beach,2383,4,beaches,Hikkaduwa,3,0
-Bundala National Park,595,4.5,nature & wildlife,Weligatta,4,40
-Kalametiya Lagoon Bird Sanctuary & Wetland Park,380,5,nature & wildlife,Tangalle,3,30
-Sinharaja Forest Reserve,612,4.5,nature & wildlife,Deniyaya,4,30
-Jungle Beach,2824,3.5,beaches,Unawatuna,1,0
+Title,Reviews,Rating,Category,Location,Duration,Budget,lati,long
+Galle Fort,3516,4.5,point of interests,Galle,3,0,6.0329,80.2168
+Bentota Beach,2305,4.5,beaches,Bentota,1,0,6.4189,80.006
+Old Town of Galle and its Fortifications,3511,4.5,point of interests,Galle,1,0,6.0329,80.2168
+Mirissa Beach,3095,4.5,beaches,Mirissa,1,0,5.9483,80.4716
+Sea Turtle Farm Galle Mahamodara,1076,4.5,nature & wildlife,Galle,2,30,6.0329,80.2168
+Hikkaduwa Beach,2383,4,beaches,Hikkaduwa,3,0,6.1395,80.1063
+Bundala National Park,595,4.5,nature & wildlife,Weligatta,4,40,6.1969,81.2409
+Kalametiya Lagoon Bird Sanctuary & Wetland Park,380,5,nature & wildlife,Tangalle,3,30,6.0243,80.7941
+Sinharaja Forest Reserve,612,4.5,nature & wildlife,Deniyaya,4,30,6.3425,80.5597
+Jungle Beach,2824,3.5,beaches,Unawatuna,1,0,6.0174,80.2489

model/src/location.py

@@ -1,27 +1,33 @@
-#import libraries
 import pandas as pd
 import json
 import sys
 from sklearn.feature_extraction.text import TfidfVectorizer
-from sklearn.metrics.pairwise import cosine_similarity
 from sklearn.metrics.pairwise import linear_kernel
+from math import radians, cos, sin, asin, sqrt
 
-#laod the dataset
+# Assuming df is your DataFrame and it has been defined and loaded with data
 df = pd.read_csv('../model/data/southern.csv')
 
 # Combine the 'Location', 'Category', 'Budget', and 'Duration' columns into a single 'Features' column
 df['Features'] = df['Location'] + ' ' + df['Category'] + ' ' + df['Budget'].astype(str) + ' ' + df['Duration'].astype(str)
 
-# Calculate TF-IDF matrix
 tfidf = TfidfVectorizer(stop_words='english')
-df['Features'] = df['Features'].fillna('')
 tfidf_matrix = tfidf.fit_transform(df['Features'])
 
-# Calculate cosine similarity matrix
-cosine_sim = linear_kernel(tfidf_matrix, tfidf_matrix)
+# Function to calculate the distance between two points using their coordinates
+def haversine(lon1, lat1, lon2, lat2):
+    lon1, lat1, lon2, lat2 = map(radians, [lon1, lat1, lon2, lat2])
+    dlon = lon2 - lon1
+    dlat = lat2 - lat1
+    a = sin(dlat/2)**2 + cos(lat1) * cos(lat2) * sin(dlon/2)**2
+    c = 2 * asin(sqrt(a))
+    r = 6371  # Radius of earth in kilometers
+    return c * r
 
 # Function to recommend locations based on similarity score
-def recommend_locations(user_input_location, user_input_categories, user_input_budget, user_input_duration, cosine_sim=cosine_sim):
+def recommend_locations(user_input_location, user_input_categories, user_input_budget, user_input_days):
+    user_input_duration = user_input_days * 8  # Calculate total duration based on number of days
+
     # Create a new dataframe to store the user's input
     user_df = pd.DataFrame([[user_input_location, ' '.join(user_input_categories), user_input_budget, user_input_duration]], columns=['Location', 'Category', 'Budget', 'Duration'])
     user_df['Features'] = user_df['Location'] + ' ' + user_df['Category'] + ' ' + user_df['Budget'].astype(str) + ' ' + user_df['Duration'].astype(str)
@@ -30,26 +36,72 @@ def recommend_locations(user_input_location, user_input_categories, user_input_b
     # Calculate the cosine similarity between the user's input and the locations in the dataframe
     cosine_sim_user = linear_kernel(user_tfidf, tfidf_matrix)
 
-    # Get the top 10 most similar locations
+    # Get the most similar locations
     sim_scores = list(enumerate(cosine_sim_user[0]))
     sim_scores = sorted(sim_scores, key=lambda x: x[1], reverse=True)
-    sim_scores = sim_scores[0:10]
-    location_indices = [i[0] for i in sim_scores]
-    recommended_locations = df.iloc[location_indices]
-    return recommended_locations
+
+    # Filter locations based on duration, location, categories, and budget
+    location_indices = [i[0] for i in sim_scores if df.iloc[i[0]]['Duration'] <= user_input_duration and df.iloc[i[0]]['Location'] == user_input_location and any(category in df.iloc[i[0]]['Category'] for category in user_input_categories) and df.iloc[i[0]]['Budget'] <= user_input_budget]
+
+    # If there are not enough locations in the input location, get other locations
+    if len(location_indices) < user_input_duration / 8:
+        remaining_duration = user_input_duration - sum(df.iloc[i]['Duration'] for i in location_indices)
+        other_location_indices = [i[0] for i in sim_scores if df.iloc[i[0]]['Duration'] <= remaining_duration and df.iloc[i[0]]['Location'] != user_input_location and any(category in df.iloc[i[0]]['Category'] for category in user_input_categories) and df.iloc[i[0]]['Budget'] <= user_input_budget]
+        for index in other_location_indices:
+            location_duration = df.iloc[index]['Duration']
+            if remaining_duration - location_duration >= 0:
+                location_indices.append(index)
+                remaining_duration -= location_duration
+            else:
+                break
+
+    # Get the coordinates of the input location
+    input_location_coordinates = df[df['Location'] == user_input_location][['lati', 'long']].values[0]
+
+    # Calculate the distance from the input location to each location in the DataFrame
+    df['Distance'] = df.apply(lambda row: haversine(input_location_coordinates[1], input_location_coordinates[0], row['long'], row['lati']), axis=1)
+
+    # Sort the locations by distance
+    location_indices = sorted(location_indices, key=lambda x: df.iloc[x]['Distance'])
+
+    # Split locations into days based on a maximum of 8 hours per day
+    recommended_locations_per_day = []
+    current_day_duration = 0
+    current_day_locations = []
+    for index in location_indices:
+        location_duration = df.iloc[index]['Duration']
+        if current_day_duration + location_duration > 8:
+            if current_day_locations:
+                recommended_locations_per_day.append(df.iloc[current_day_locations])
+            current_day_duration = location_duration
+            current_day_locations = [index]
+        else:
+            current_day_locations.append(index)
+            current_day_duration += location_duration
+        if len(recommended_locations_per_day) == user_input_days:
+            break
+    if current_day_locations and len(recommended_locations_per_day) < user_input_days:
+        recommended_locations_per_day.append(df.iloc[current_day_locations])
+
+    return recommended_locations_per_day
 
 # Get user input
 user_input_location = "Tangalle"
 user_input_category = ["point of interests", "beaches", "nature & wildlife"]
-user_input_budget = 1100
-user_input_duration = 7
+user_input_budget = 1000
+user_input_days = 3  # Number of days
 
 # Recommend locations based on the user's input
-recommendations = recommend_locations(user_input_location, user_input_category, user_input_budget, user_input_duration)
-
-# print the results as a JSON string
-print(json.dumps(
-    {
-        "location": recommendations.to_dict('records')
-    }
-))
+recommendations = recommend_locations(user_input_location, user_input_category, user_input_budget, user_input_days)
+
+recommendations_list = []
+for i, day in enumerate(recommendations):
+    recommendations_list.append({
+        "Day": i + 1,
+        "Locations": day.to_dict('records')
+        })
+
+# Print the list as a JSON array
+print(json.dumps(recommendations_list))
+
+

model/src/util/recommendation system.ipynb

@@ -9,12 +9,7 @@
     "#import libraries\n",
     "import pandas as pd\n",
     "import matplotlib.pyplot as plt\n",
-    "import seaborn as sns\n",
-    "import sys\n",
-    "import json\n",
-    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
-    "from sklearn.metrics.pairwise import linear_kernel\n",
-    "from sklearn.metrics.pairwise import cosine_similarity"
+    "import seaborn as sns"
    ]
   },
   {

server/index.js

@@ -17,7 +17,7 @@ app.use('/', stayRoute)
 app.use('/', locationRoute)
 
 app.listen(process.env.PORT, () => {
-    console.log(`app is listening on port ${process.env.PORT}!`)
+    console.log(`server is listening on port ${process.env.PORT}!`)
 })
 
 module.exports = app;

server/routes/locationRoute.js

@@ -24,8 +24,7 @@ router.get('/api/location/', (req, res) => {
     python.on('close', (code) => {
         console.log(`child process close all stdio with code ${code}`);
         // send data to browser
-        res.send(dataToSend)
-
+        res.send(dataToSend); 
     });
 })
 

server/routes/stayRoute.js

@@ -25,7 +25,6 @@ router.get('/api/stay/', (req, res) => {
         console.log(`child process close all stdio with code ${code}`);
         // send data to browser
         res.send(dataToSend)
-
     });
 })