def mostFrecuencyImputingHandler(self, request):
print("mostFrecuencyImputingHandler")
df = dataframeHandler.getDataframe()
column = request.form.get('column')
df[[column]] = df[[column]].fillna(df[[column]].mode().iloc[0])
pd.set_option("max_columns", None) # show all cols
dataframeHandler.saveDataframe(df)
def kNearestNeighborsImputingHandler(self, request):
df = dataframeHandler.getDataframe()
column = request.form.get('column')
n_neighbors = request.form.get('n_neighbors')
encoders = dict()
df_copy = df.copy()
for col_name in df_copy.columns:
print(col_name, df_copy[col_name].dtype)
if (df_copy[col_name].dtype == "object"):
series = df_copy[col_name]
label_encoder = preprocessing.LabelEncoder()
df_copy[col_name] = pd.Series(
label_encoder.fit_transform(series[series.notnull()]),
index=series[series.notnull()].index
)
encoders[col_name] = label_encoder
else:
print("Column", col_name, "not encoded")
imputer = KNNImputer(n_neighbors=int(n_neighbors))
imputedData = imputer.fit_transform(df_copy)
imputed_dataframe = pd.DataFrame(imputedData, columns=df_copy.columns)
if(df[column].dtype == "object"):
imputed_dataframe[[column]] = imputed_dataframe[[column]].astype(int)
imputed_dataframe[[column]] = encoders[column].inverse_transform(imputed_dataframe[[column]])
df[[column]] = imputed_dataframe[[column]]
dataframeHandler.saveDataframe(df)
def averageImputingHandler(self, request):
df = dataframeHandler.getDataframe()
column = request.form.get('column')
axis = request.form.get('axis')
print("axis: ", axis)
print("column: ", column)
df[[column]] = df[[column]].fillna(df.mean(axis=int(axis)))
pd.set_option("max_columns", None) # show all cols
dataframeHandler.saveDataframe(df)
def setTypeHandler(self, request):
_type = request.form.get('type')
if _type in self.typesHandlers:
self.typesHandlers[_type](request)
else:
column = request.form.get('column')
df = dataframeHandler.getDataframe()
df[[column]] = df[[column]].astype(_type)
dataframeHandler.saveDataframe(df)
def interpolationImputingHandler(self, request):
df = dataframeHandler.getDataframe()
column = request.form.get('column')
method = request.form.get('method')
order = request.form.get('order')
axis = request.form.get('axis')
# df = df.interpolate(method='polynomial', order=2, axis=0)
df[[column]] = df[[column]].interpolate(method=method, order=int(order), axis=int(axis))
pd.set_option("max_columns", None) # show all cols
dataframeHandler.saveDataframe(df)
def updateCellHandler(self, request):
column = request.form.get('column')
row = int(request.form.get('row'))
value = request.form.get('value')
if not value or value.isspace():
value = None
df = dataframeHandler.getDataframe()
pd.set_option("max_columns", None) # show all cols
df.at[row, column] = value
dataframeHandler.saveDataframe(df)
def exampleActionHandler(self, request):
df = dataframeHandler.getDataframe()
column = request.form.get('column')
axis = request.form.get('axis')
# ** HERE YOUR CODE FOR EXAMPLE ACTION HANDLER OF THIS PLUGIN **
# modify df and it will be saved with dataframeHandler class in the
# local cache and then returned in
# Obtain the params from the request
exampleSelect = request.form.get('exampleSelect')
exampleNumber = request.form.get('exampleNumber')
exampleString = request.form.get('exampleString')
exampleFile = request.files['exampleFile']
# do something like print params
print("exampleSelect: ", exampleSelect)
print("exampleNumber: ", exampleNumber)
print("exampleString: ", exampleString)
print("exampleFile: ", exampleFile)
# always save the dataframe in the local cache
dataframeHandler.saveDataframe(df)
def setTypeFloatHandler(self, request):
column = request.form.get('column')
df = dataframeHandler.getDataframe()
df[column] = pd.to_numeric(df[column], errors='coerce')
dataframeHandler.saveDataframe(df)
def deleteColumnHandler(self, request):
column = request.form.get('column')
df = dataframeHandler.getDataframe()
df.drop(column, axis=1, inplace=True)
dataframeHandler.saveDataframe(df)
def deleteRowHandler(self, request):
row = int(request.form.get('row'))
df = dataframeHandler.getDataframe()
df = df.drop(row)
df = df.reset_index(drop=True)
dataframeHandler.saveDataframe(df)
def csvHandler (self, file: any):
df = pd.read_csv(file)
dataframeHandler.saveDataframe(df)