Skip to content

Commit

Permalink
Merge pull request #364 from Palashio/dashboard-issue
Browse files Browse the repository at this point in the history 
Text classification fixed max features issues
  • Loading branch information
@anas-awadalla
anas-awadalla authored Sep 22, 2020
2 parents 7d95c80 + 1104323 commit cc2a60b
Show file tree
Hide file tree
Showing 5 changed files with 126 additions and 115 deletions.
205 changes: 107 additions & 98 deletions libra/dashboard/LibEDA.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@
import seaborn as sns
import streamlit as st
import sys

sys.path.insert(1, '.')
from libra.query.dimensionality_red_queries import dimensionality_RF, dimensionality_PCA, dimensionality_ICA
from libra import client
Expand Down Expand Up @@ -37,87 +38,86 @@ def get_csv_download_link(self, df):

def visualize_bar(self, df, x_axis, y_axis, legend, tooltips):
graph = alt.Chart(df).mark_bar().encode(
x = x_axis,
y = y_axis,
color = legend
).interactive().properties(width = self.width, height = self.height)
x=x_axis,
y=y_axis,
color=legend
).interactive().properties(width=self.width, height=self.height)
st.text("")
st.text("")
st.write(graph)


def visualize_circle(self, df, x_axis, y_axis, legend):
graph = alt.Chart(df).mark_circle(size = 60).encode(
x = x_axis,
y = y_axis,
color = legend
).interactive().properties(width = self.width, height = self.height)
graph = alt.Chart(df).mark_circle(size=60).encode(
x=x_axis,
y=y_axis,
color=legend
).interactive().properties(width=self.width, height=self.height)
st.text("")
st.text("")
st.write(graph)

def visualize_line(self, df, x_axis, y_axis, legend):
graph = alt.Chart(df).mark_line().encode(
x = x_axis,
y = y_axis,
color = legend
).interactive().properties(width = self.width, height = self.height)
x=x_axis,
y=y_axis,
color=legend
).interactive().properties(width=self.width, height=self.height)
st.text("")
st.text("")
st.write(graph)

def visualize_area(self, df, x_axis, y_axis, legend):
graph = alt.Chart(df).mark_area().encode(
x = x_axis,
y = y_axis,
color = legend
).interactive().properties(width = self.width, height = self.height)
x=x_axis,
y=y_axis,
color=legend
).interactive().properties(width=self.width, height=self.height)
st.text("")
st.text("")
st.write(graph)

def visualize_box(self, df, x_axis, y_axis, legend, tooltips):
graph = alt.Chart(df).mark_boxplot().encode(
x = x_axis,
y = y_axis,
color = legend
).interactive().properties(width = self.width, height = self.height)
x=x_axis,
y=y_axis,
color=legend
).interactive().properties(width=self.width, height=self.height)
st.text("")
st.text("")
st.write(graph)

def visualize_count(self, df, x_axis):
graph = alt.Chart(df).mark_bar().encode(
x = x_axis,
y = 'count('+x_axis+'):Q',
color = x_axis+':N',
).interactive().properties(width = self.width, height = self.height)
x=x_axis,
y='count(' + x_axis + '):Q',
color=x_axis + ':N',
).interactive().properties(width=self.width, height=self.height)
st.text("")
st.text("")
st.write(graph)

def visualize_heatmap(self, df):
corrMatrix = df.corr().reset_index().melt('index')
corrMatrix.columns = ['X','Y','Correlation']
corrMatrix.columns = ['X', 'Y', 'Correlation']

base = alt.Chart(corrMatrix).transform_filter(
alt.datum.X < alt.datum.Y).encode(
x = 'X',
y = 'Y',
).properties(
width = self.width,
height = self.height
)
x='X',
y='Y',
).properties(
width=self.width,
height=self.height
)

rects = base.mark_rect().encode(
color = 'Correlation'
color='Correlation'
)

text = base.mark_text(
size = 30
size=30
).encode(
text = alt.Text('Correlation', format = '.2f'),
color = alt.condition(
text=alt.Text('Correlation', format='.2f'),
color=alt.condition(
"datum.Correlation > 0.5",
alt.value('white'),
alt.value('black')
Expand All @@ -130,11 +130,11 @@ def visualize_heatmap(self, df):
def visualize_selection(self, df, x_axis, y_axis, legend):
brush = alt.selection_interval()
graph = alt.Chart(df).mark_point().encode(
x = x_axis+':Q',
y = y_axis+':Q',
color = alt.condition(brush, legend+':N', alt.value('lightgray'))
#tooltip = tooltips
).properties(width = self.width, height = self.height).add_selection(
x=x_axis + ':Q',
y=y_axis + ':Q',
color=alt.condition(brush, legend + ':N', alt.value('lightgray'))
# tooltip = tooltips
).properties(width=self.width, height=self.height).add_selection(
brush
)
st.text("")
Expand All @@ -150,21 +150,20 @@ def visualize_distribution(self, df, x_axis):
except:
st.subheader("Cannot Build Distribution Plot")


def main(self):
st.markdown(self.hide_menu_style, unsafe_allow_html=True)

st.sidebar.title('Sections')
page = st.sidebar.radio("Go To Page",['Homepage', 'Data View', 'EDA','Dimensionality Reduction'])
#data = st.file_uploader('Upload Your Dataset', type = 'csv')
page = st.sidebar.radio("Go To Page", ['Homepage', 'Data View', 'EDA', 'Dimensionality Reduction'])
# data = st.file_uploader('Upload Your Dataset', type = 'csv')
df = st.cache(pd.read_csv)(sys.argv[1])

if page == 'Homepage':
st.title("Welcome To The LibEDA App!")
st.write("The Purpose Of This App Is To Simplify The EDA Process With Just The Click Of The Mouse.")
st.write("The LibEDA App Has A Lot Of Features That Can Be Used For EDA.")

st.subheader("View Your Original Data")
st.subheader("View Your Original Data")
st.write("")
st.write("What Can You Do With Your Data View?")
st.write("1. View Entire Data")
Expand All @@ -173,8 +172,8 @@ def main(self):
st.write("4. View Single Column")
st.write("5. View Data Summary")
st.write("")
st.write("")
st.write("")

st.subheader("You Can Vizualize Your Data With Just A Click.")
st.write("")
st.write("The Different Types Of Plots You Can Generate Are - ")
Expand Down Expand Up @@ -206,29 +205,30 @@ def main(self):
elif page == 'Data View':
st.title("Data View")
st.subheader("Here You Can Take Different Views Of Your Data")
display = st.selectbox("Choose What You What Data You Want To View",
['View Entire Data', 'View First N Rows Of Data', 'Show Dimensions of Data', 'View Single Column', 'View Data Summary'],
index = 0)
display = st.selectbox("Choose What You What Data You Want To View",
['View Entire Data', 'View First N Rows Of Data', 'Show Dimensions of Data',
'View Single Column', 'View Data Summary'],
index=0)

if display == 'View Entire Data':
st.write(df)

elif display == 'View First N Rows Of Data':
rows = st.slider('How Many Rows Do You Want To See?',5,50)
rows = st.slider('How Many Rows Do You Want To See?', 5, 50)
st.write("Your Data is Displayed Below")
st.write(df.head(rows))

elif display == 'Show Dimensions of Data':
data_dim = st.radio("What Dimension Do You Want to Show", ("Rows", "Columns"))
if data_dim == "Rows":
st.text("Showing Number of Rows")
st.write(len(df))
if data_dim == "Columns":
st.text("Showing Number of Columns")
st.write(df.shape[1])
data_dim = st.radio("What Dimension Do You Want to Show", ("Rows", "Columns"))
if data_dim == "Rows":
st.text("Showing Number of Rows")
st.write(len(df))
if data_dim == "Columns":
st.text("Showing Number of Columns")
st.write(df.shape[1])

elif display == 'View Single Column':
col = st.selectbox("Select Column You Want To View", df.columns, index = 0)
col = st.selectbox("Select Column You Want To View", df.columns, index=0)
st.write(df[col])

elif display == 'View Data Summary':
Expand All @@ -238,75 +238,82 @@ def main(self):
st.sidebar.title('What To Do?')
st.sidebar.info("You Can Choose Your Inputs In The Dropdown For Which You Want The Plot For")
st.sidebar.title('Note:')
st.sidebar.info("Some Plots Are Interactive. You Can Choose Your Tooltips and Hover Over The Plot To View The Selected Tooltips.")
st.sidebar.info(
"Some Plots Are Interactive. You Can Choose Your Tooltips and Hover Over The Plot To View The Selected Tooltips.")

st.title('Exploratory Data Analysis')
plot_types = ['Barplot','Scatterplot','Lineplot','Areaplot','Boxplot','Countplot','Correlation Heatmap','Selection Plot', 'Distribution Plot']
type_of_plot = st.selectbox("Choose Type Of Plot", plot_types,index = 0)
plot_types = ['Barplot', 'Scatterplot', 'Lineplot', 'Areaplot', 'Boxplot', 'Countplot',
'Correlation Heatmap', 'Selection Plot', 'Distribution Plot']
type_of_plot = st.selectbox("Choose Type Of Plot", plot_types, index=0)

if type_of_plot == 'Barplot':
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index = len(df.columns)-1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index = len(df.columns)-2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index = len(df.columns)-3)
tooltips = st.multiselect("Choose Variable(s) For Tooltips", list(df.columns), default = [list(df.columns)[0]])
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index=len(df.columns) - 1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index=len(df.columns) - 2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index=len(df.columns) - 3)
tooltips = st.multiselect("Choose Variable(s) For Tooltips", list(df.columns),
default=[list(df.columns)[0]])
self.visualize_bar(df, x_axis, y_axis, legend, tooltips)

elif type_of_plot == 'Scatterplot':
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index = len(df.columns)-1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index = len(df.columns)-2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index = len(df.columns)-3)
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index=len(df.columns) - 1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index=len(df.columns) - 2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index=len(df.columns) - 3)
self.visualize_circle(df, x_axis, y_axis, legend)

elif type_of_plot == 'Lineplot':
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index = len(df.columns)-1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index = len(df.columns)-2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index = len(df.columns)-3)
tooltips = st.multiselect("Choose Variable(s) For Tooltips", list(df.columns), default = [list(df.columns)[0]])
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index=len(df.columns) - 1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index=len(df.columns) - 2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index=len(df.columns) - 3)
tooltips = st.multiselect("Choose Variable(s) For Tooltips", list(df.columns),
default=[list(df.columns)[0]])
self.visualize_line(df, x_axis, y_axis, legend, tooltips)

elif type_of_plot == 'Areaplot':
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index = len(df.columns)-1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index = len(df.columns)-2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index = len(df.columns)-3)
tooltips = st.multiselect("Choose Variable(s) For Tooltips", list(df.columns), default = [list(df.columns)[0]])
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index=len(df.columns) - 1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index=len(df.columns) - 2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index=len(df.columns) - 3)
tooltips = st.multiselect("Choose Variable(s) For Tooltips", list(df.columns),
default=[list(df.columns)[0]])
self.visualize_area(df, x_axis, y_axis, legend, tooltips)

elif type_of_plot == 'Boxplot':
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index = len(df.columns)-1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index = len(df.columns)-2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index = len(df.columns)-3)
tooltips = st.multiselect("Choose Variable(s) For Tooltips", list(df.columns), default = [list(df.columns)[0]])
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index=len(df.columns) - 1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index=len(df.columns) - 2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index=len(df.columns) - 3)
tooltips = st.multiselect("Choose Variable(s) For Tooltips", list(df.columns),
default=[list(df.columns)[0]])
self.visualize_box(df, x_axis, y_axis, legend, tooltips)

elif type_of_plot == 'Countplot':
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index = len(df.columns)-1)
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index=len(df.columns) - 1)
self.visualize_count(df, x_axis)

elif type_of_plot == 'Correlation Heatmap':
st.text("Below Is The Heatmap That Indicates The Correlation Amongst The Data Columns")
self.visualize_heatmap(df)

elif type_of_plot == 'Selection Plot':
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index = len(df.columns)-1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index = len(df.columns)-2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index = len(df.columns)-3)
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index=len(df.columns) - 1)
y_axis = st.selectbox("Choose A Variable For Y-Axis", df.columns, index=len(df.columns) - 2)
legend = st.selectbox("Choose A Variable For The Legend", df.columns, index=len(df.columns) - 3)
self.visualize_selection(df, x_axis, y_axis, legend)

elif type_of_plot == 'Distribution Plot':
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index = len(df.columns)-1)
x_axis = st.selectbox("Choose A Variable For X-Axis", df.columns, index=len(df.columns) - 1)
self.visualize_distribution(df, x_axis)

elif page == 'Dimensionality Reduction':
st.sidebar.title('What To Do?')
st.sidebar.info("You Can Choose What Technique You Want To Implement From The Dropdown. Post Which, You Can Input The Parameters In The Text Box.\
Once The Transformation Is Completed, You Will Get A Download Link With Which You Can Download The Transformed Dataset.")
st.sidebar.title('Note:')
st.sidebar.info("Transformations Will Take Time. Changing Sections During A Task Will Result In Loss Of Progress.")
st.sidebar.info(
"Transformations Will Take Time. Changing Sections During A Task Will Result In Loss Of Progress.")
st.title("Dimensionality Reduction")
st.subheader("Here You Can View The Transformed Data")
display = st.selectbox("Choose The Dimensionality Reduction Technique",
['Random Forest', 'PCA', 'ICA'],
index = 0)
display = st.selectbox("Choose The Dimensionality Reduction Technique",
['Random Forest', 'PCA', 'ICA'],
index=0)

if display == 'Random Forest':
instruction = st.text_input("Enter Your Instruction")
Expand All @@ -320,7 +327,7 @@ def main(self):
st.write("")
if st.button("Transform"):
st.write("Transforming...")
out = dimensionality_RF(instruction, sys.argv[1], target = target, y = y, n_features = int(n_features))
out = dimensionality_RF(instruction, sys.argv[1], target=target, y=y, n_features=int(n_features))
st.write("")
st.write("Transformed Data")
st.write("")
Expand All @@ -345,7 +352,8 @@ def main(self):
st.write("")
if st.button("Transform"):
st.write("Transforming...")
out = dimensionality_PCA(instruction, sys.argv[1], ca_threshold=(lambda x:None if x=="" else int(x)(ca_thresh)) )
out = dimensionality_PCA(instruction, sys.argv[1],
ca_threshold=(lambda x: None if x == "" else int(x)(ca_thresh)))
st.write("")
st.write("Transformed Data")
st.write("")
Expand All @@ -371,7 +379,7 @@ def main(self):
st.write("")
if st.button("Transform"):
st.write("Transforming...")
out = dimensionality_ICA(instruction, sys.argv[1], target = target, y = y)
out = dimensionality_ICA(instruction, sys.argv[1], target=target, y=y)
st.write("")
st.write("Transformed Data")
st.write("")
Expand All @@ -394,5 +402,6 @@ def main(self):
predict_what = st.text_input("What Is Your Query?", '')
'''


eda_call = edaDashboardback()
eda_call.main()
Loading

0 comments on commit cc2a60b

Please sign in to comment.