def loadData() :
# Load Data
sql = "SELECT user_handle,course_id,author_handle,level,sum(view_time_seconds) view_time_seconds from user_course_views group by user_handle,course_id,author_handle,level"
data = db.loadDatabySQL(sql)
data.head()
# generate sequence numbers for string column values
data["course_id"] = le.fit_transform(data["course_id"])
data["level"] = le.fit_transform(data["level"])
# Explore Data
#df = pd.DataFrame(data)
print ("head : ")
print ("------")
print (data[1:5])
print("")
print ("shape ")
print ("------")
print (data.shape)
print("")
print ("Describe Result" )
print ("---------------")
print (data.describe())
print("")
# Prepare X and Y
X = data[[ "course_id","level"]]
y = data["view_time_seconds"]
#Perform linear regression using formula api
df2=pd.DataFrame(X,columns=['course_id','level'])
df2['view_time_seconds']=pd.Series(y)
return(df2)
def predict_CL():
print(
"Running - IRIS_StudentClassification_KNN_1_IRIS_StudentClassification_KNN_1_CLCL.py"
)
# Load Data
sql = "SELECT user_handle,course_id,author_handle,level,sum(view_time_seconds) view_time_seconds from user_course_views group by user_handle,course_id,author_handle,level"
df = db.loadDatabySQL(sql)
df.head()
# Remove header row
data = df.iloc[1:]
# Remove last row as it contains the column header in string format
data = data.drop(data.index[len(data) - 1])
data["course_id"] = le.fit_transform(data["course_id"])
data["level"] = le.fit_transform(data["level"])
data = data.apply(pd.to_numeric)
# Prepare X and Y
iris_X = data[['author_handle', 'course_id']]
iris_y = data['level']
# consider 400 rows for taining set and 100 rows of data for test set
iris_X_train = iris_X[:-400]
iris_y_train = iris_y[:-400]
iris_X_test = iris_X[-100:]
iris_y_test = iris_y[-100:]
# prepare the KNN classifier model
from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn.fit(iris_X_train, iris_y_train)
KNeighborsClassifier(algorithm='auto',
leaf_size=30,
metric='minkowski',
metric_params=None,
n_jobs=1,
n_neighbors=5,
p=2,
weights='uniform')
# predict using test set
pr = knn.predict(iris_X_test)
#print(pr)
# display the results
dfFinal = pd.DataFrame(iris_X_test)
dfFinal['Actual result'] = iris_y_test
dfFinal['Predicted Result'] = pr
# Return first 10 rows for actual and predicted set of rows
df1 = dfFinal.head()
htmltext = df1.to_html()
print(df1)
return (htmltext)
def loadData() :
# Read Data
df = db.loadData("user_assessment_scores")
df = df.drop(['user_assessment_date'], 1)
# Remove header row
df = df.iloc[1:]
df["assessment_tag"] = le.fit_transform(df["assessment_tag"])
df.head()
df = df.drop(['user_handle'], 1)
return(df)
def loadData():
df = db.loadData("user_interests")
df.head()
# Remove header row
df = df.iloc[1:]
# drop unwanted columns
df = df.drop(['date_followed'], 1)
# convert interest_tag to serial numbers
df["interest_tag"] = le.fit_transform(df["interest_tag"])
return (df)
def loadData():
# load data
df = db.loadData("user_course_views")
# select only first 5 columns
df = df.iloc[:, 0:5]
df = df.drop(['view_date', 'author_handle'], 1)
# Convert the string data to numeric series
df["course_id"] = le.fit_transform(df["course_id"])
df["level"] = le.fit_transform(df["level"])
# Remove header row
df = df.iloc[1:]
return (df)
def loadData():
df = db.loadData("user_course_views")
df.head()
# select only first 5 columns
df = df.iloc[:, 0:5]
# drop unwanted columns for this analysis
df = df.drop(['view_date', 'user_handle'], 1)
# Convert the string data to numeric series
df["course_id"] = le.fit_transform(df["course_id"])
df["level"] = le.fit_transform(df["level"])
# Remove header row
df = df.iloc[1:]
return (df)
def loadData() :
print("Running - IRIS_StudentClassification_Views__AC_5.py")
#Read Data
df = db.loadData("user_course_views")
# select only first 5 columns
df = df.iloc[:,0:5]
# remove unwanted columns for this analysis
df = df.drop(['view_date','user_handle'], 1)
# Convert the string data to numeric series
df["course_id"] = le.fit_transform(df["course_id"])
df["level"] = le.fit_transform(df["level"])
# Remove header row
df = df.iloc[1:]
return(df)
def loadData():
print(
"Running - IRIS_StudentClassification_KNN_1_IRIS_StudentClassification_KNN_1_CLCL.py"
)
# Load Data
df = db.loadData("user_assessment_scores")
# remove unwanted columns
df = df.drop(['user_assessment_date'], 1)
# conver the column values to sequence numbers
df["assessment_tag"] = le.fit_transform(df["assessment_tag"])
df = df.drop(['assessment_tag'], 1)
# Remove header row
df = df.iloc[1:]
return (df)
def loadData():
#Read Data
df = db.loadData("user_course_views")
df.head()
# select only first 5 columns
df = df.iloc[:, 0:5]
# remove author handle and view date columns
df = df.drop(['view_date', 'author_handle'], 1)
df.head()
# Convert the string data to numeric series
df["course_id"] = le.fit_transform(df["course_id"])
df["level"] = le.fit_transform(df["level"])
# remove duplicates
df = df.drop_duplicates(subset=['user_handle', 'course_id', 'level'],
keep=False)
return (df)
# Remove header row
df = df.iloc[1:]