feature: adding streamlit interface to wise pizza

requirements.txt

@@ -9,3 +9,4 @@ scipy>=1.8.0
 tqdm
 cloudpickle
 pivottablejs
+streamlit==1.28.0

streamlit/streamlit_app.py

@@ -0,0 +1,289 @@
+"""An example of Streamlit leveraging Wise pizza."""
+
+import streamlit as st
+import os, sys
+import pandas as pd
+
+import warnings
+warnings.filterwarnings("ignore")
+
+root_path = os.path.realpath('../..')
+
+# this assumes that all of the following files are checked in the same directory
+sys.path.append(os.path.join(root_path,"wise-pizza"))
+
+# create data-related directories
+data_dir = os.path.realpath(os.path.join(root_path, 'wise-pizza/data'))
+if not os.path.isdir(data_dir):
+    os.mkdir(data_dir)
+
+from wise_pizza import explain_levels, explain_changes_in_totals, explain_changes_in_average
+
+# False if you want nice interactive plots
+# True if you want static plots (Doesn't work on all platforms yet)
+plot_is_static = False
+
+
+# SETTING PAGE CONFIG TO WIDE MODE AND ADDING A TITLE AND FAVICON
+st.set_page_config(layout="wide", page_title="Wise Pizza", page_icon=":pizza:")
+
+st.title('Wise Pizza powered by Streamlit')
+st.text('Only categorical columns are accepted, bucket the numeric ones if you wanna use those')
+
+# upload the file from the computer
+def load_data_upload():
+    uploaded_file = st.file_uploader("Choose a file")
+    if not uploaded_file:
+        st.warning('Please input a dataset.')
+        st.stop()
+    st.success('Dataset inputted.')
+    data = pd.read_csv(uploaded_file)
+    return data
+
+on = st.toggle('Use sample data from Github')
+url_data = (r'https://raw.githubusercontent.com/transferwise/wise-pizza/main/data/synth_data.csv')
+
+# select the datasource, either local or from github
+if on:
+    st.write(f'Downloading data from {url_data}!')
+    df = pd.read_csv(url_data)
+else:
+    df=load_data_upload()
+
+# show dataset preview
+st.text('Table preview')
+st.table(df.head(10))
+
+# ask the user via streamlit for the target column
+totals = st.selectbox(
+   "What is the target column that you want to analyse? e.g. GMV/revenue",
+   df.columns,
+   index=None,
+   placeholder="Select target column",
+)
+st.write('You selected:', totals)
+
+# ask the user via streamlit for the size column
+size = st.selectbox(
+   "What is the volume column of your dataset? e.g. number of users/transactions",
+   df.columns,
+   index=None,
+   placeholder="Select volume column",
+)
+st.write('You selected:', size)
+
+# ask the user via streamlit for the flag column
+flag_column = st.selectbox(
+   "What is the flag column of your dataset you wanan split it by? Ensure this column is binary",
+   df.columns,
+   index=None,
+   placeholder="Select time column",
+)
+st.write('You selected:', flag_column)
+
+# wait until flag column is added
+if not flag_column:
+        st.stop()
+
+# calculate unique flags in the dataset
+flags = sorted(df[flag_column].unique())  
+
+# show an error message if the specified flag column is not binary
+if len(flags)>2:
+    st.error('Your flag is not binary', icon="🚨")
+
+# allow users to define what's "old" and "new" in the comparison
+flags_option = st.selectbox(
+    'Which one in your data belongs to group A?',
+    (flags))
+
+# listing all potential dimensions in the dataframe. 
+# all of them are potential columns to exclude
+candidates_excluded_columns = [element for element in df.columns if element not in [totals,size,flag_column]]
+
+# list with specified columns to exclude.
+excluded_columns = st.multiselect(
+    'Please select all columns that you want to exclude from the analysis',
+    candidates_excluded_columns)
+
+# all non dimensional columns are the ones picked by the user (if any), plus the ones that indicate totals, size and flag
+non_dimensional_columns = excluded_columns + [totals,size,flag_column]
+
+# calculating dimensions
+dims = [element for element in df.columns if element not in non_dimensional_columns]
+
+# creating the dataframes for the comparison calculations
+data = df[df[flag_column] != flags_option]  # take the group to compare to
+pre_data = df[df[flag_column] == flags_option]  # take the group to be compared
+
+st.subheader('Finding the juiciest slices', divider='rainbow')
+st.text('This section does not compare groups, but rather checks which features have the most impact in the target column you selected.')
+
+## finding juiciest slices
+sf = explain_levels(
+    df=df,
+    dims=dims,
+    total_name=totals,
+    size_name=size,
+    max_depth=2,
+    min_segments=20,
+    solver="lasso",
+    return_fig=True
+)
+
+# storing the plot in a variable
+plot_sf=sf.plot(width=500, height=500)
+# exposing the plot via streamlit
+st.plotly_chart(plot_sf, use_container_width=True)
+
+st.subheader('Analysing differences', divider='rainbow')
+st.text('This section does compare the two groups defined by the flag. Old total is the group A you selected in the dropdown')
+
+## running explain calculations
+sf1 = explain_changes_in_totals(
+    df1=pre_data,
+    df2=data,
+    dims=dims,
+    total_name=totals,
+    size_name=size,
+    max_depth=2,
+    min_segments=20,
+    how="totals",
+    solver="lasso",
+    return_fig=True
+)
+# specifying a two column layout on streamlit
+col1, col2 = st.columns(2)
+# storing the plots in variables
+# exposing the plots via streamlit
+with col1:
+    plot_sf1=sf1.plot(width=500, height=500, plot_is_static=plot_is_static)[0]
+    st.plotly_chart(plot_sf1, use_container_width=True)
+
+with col2:
+    plot_sf2=sf1.plot(width=500, height=500, plot_is_static=plot_is_static)[1]
+    st.plotly_chart(plot_sf2, use_container_width=True)
+
+st.subheader('Decomposing differences', divider='rainbow')
+st.text('`split_fits` to separately decompose contribution of size changes and average changes')
+
+## running explain calculations
+sf2 = explain_changes_in_totals(
+    df1=pre_data,
+    df2=data,
+    dims=dims,
+    total_name=totals,
+    size_name=size,
+    max_depth=1,
+    min_segments=10,
+    how="split_fits",
+    solver="lasso",
+    return_fig=True
+)
+# storing the plot in a variable
+# exposing the plot via streamlit
+plot_sf=sf2.plot(width=500, height=500)
+st.plotly_chart(plot_sf, use_container_width=True)
+
+st.text('`extra_dim` to treat size vs average change contribution as an additional dimension')
+
+## running explain calculations
+sf3 = explain_changes_in_totals(
+    df1=pre_data,
+    df2=data,
+    dims=dims,
+    total_name=totals,
+    size_name=size,
+    max_depth=2,
+    min_segments=20,
+    how="extra_dim",
+    solver="lasso",
+    return_fig=True
+)
+# specifying a two column layout on streamlit
+col1, col2 = st.columns(2)
+
+# storing the plots in variables
+# exposing the plots via streamlit
+with col1:
+    plot_sf1=sf3.plot(width=500, height=500, plot_is_static=plot_is_static)[0]
+    st.plotly_chart(plot_sf1, use_container_width=True)
+
+with col2:
+    plot_sf2=sf3.plot(width=500, height=500, plot_is_static=plot_is_static)[1]
+    st.plotly_chart(plot_sf2, use_container_width=True)
+
+st.text('`force_dim` like extra_dim, but each segment must contain a Change_from constraint')
+
+## running explain calculations
+sf3 = explain_changes_in_totals(
+    df1=pre_data,
+    df2=data,
+    dims=dims,
+    total_name=totals,
+    size_name=size,
+    max_depth=2,
+    min_segments=15,
+    how="force_dim",
+    solver="lasso",
+    return_fig=True
+)
+# specifying a two column layout on streamlit
+col1, col2 = st.columns(2)
+
+# storing the plots in variables
+# exposing the plots via streamlit
+with col1:
+    plot_sf1=sf3.plot(width=500, height=500, plot_is_static=plot_is_static)[0]
+    st.plotly_chart(plot_sf1, use_container_width=True)
+
+with col2:
+    plot_sf2=sf3.plot(width=500, height=500, plot_is_static=plot_is_static)[1]
+    st.plotly_chart(plot_sf2, use_container_width=True)
+
+st.subheader('Explaining changes in average', divider='rainbow')
+
+## running explain calculations
+sf4 = explain_changes_in_average(
+    df1=pre_data,
+    df2=data,
+    dims=dims,
+    total_name=totals,
+    size_name=size,
+    max_depth=2,
+    min_segments=20,
+    how="totals",
+    solver="lasso",
+    return_fig=True
+)
+# specifying a two column layout on streamlit
+col1, col2 = st.columns(2)
+
+# storing the plots in variables
+# exposing the plots via streamlit
+with col1:
+    plot_sf1=sf4.plot(width=500, height=500, plot_is_static=plot_is_static)[0]
+    st.plotly_chart(plot_sf1, use_container_width=True)
+
+with col2:
+    plot_sf2=sf4.plot(width=500, height=500, plot_is_static=plot_is_static)[1]
+    st.plotly_chart(plot_sf2, use_container_width=True)
+
+## running explain calculations
+sf6 = explain_changes_in_average(
+    df1=pre_data,
+    df2=data,
+    dims=dims,
+    total_name=totals,
+    size_name=size,
+    max_depth=2,
+    min_segments=20,
+    how="split_fits",
+    solver="lasso",
+    return_fig=True
+)
+
+# storing the plot in a variable
+plot_sf=sf6.plot(width=500, height=500)
+# exposing the plot via streamlit
+st.plotly_chart(plot_sf, use_container_width=True)

wise_pizza/explain.py

@@ -26,6 +26,7 @@ def explain_changes_in_average(
     force_add_up: bool = False,
     constrain_signs: bool = True,
     verbose: int = 0,
+    return_fig: bool = False
 ):
     """
     Find segments most useful in explaining the difference between the averages of the two datasets
@@ -48,6 +49,7 @@ def explain_changes_in_average(
     @param constrain_signs: Whether to constrain weights of segments to have the same
     sign as naive segment averages
     @param verbose: If set to a truish value, lots of debug info is printed to console
+    @param return_fig: If set to true, plot returns the figure object, otherwise shows the figures
     @return: A fitted object
     """
     df1 = df1.copy()
@@ -86,6 +88,7 @@ def explain_changes_in_average(
         force_add_up=force_add_up,
         constrain_signs=constrain_signs,
         verbose=verbose,
+        return_fig=return_fig
     )
 
     if hasattr(sf, "pre_total"):
@@ -119,6 +122,7 @@ def explain_changes_in_totals(
     force_add_up: bool = False,
     constrain_signs: bool = True,
     verbose: int = 0,
+    return_fig: bool = False
 ):
     """
     Find segments most useful in explaining the difference between the totals of the two datasets
@@ -141,6 +145,7 @@ def explain_changes_in_totals(
     @param constrain_signs: Whether to constrain weights of segments to have the same
     sign as naive segment averages
     @param verbose: If set to a truish value, lots of debug info is printed to console
+    @param return_fig: If set to true, plot returns the figure object, otherwise shows the figures
     @return: A fitted object
     """
 
@@ -181,6 +186,7 @@ def explain_changes_in_totals(
             force_add_up=force_add_up,
             constrain_signs=constrain_signs,
             verbose=verbose,
+            return_fig=return_fig
         )
 
         sf_avg = explain_levels(
@@ -195,6 +201,7 @@ def explain_changes_in_totals(
             force_add_up=force_add_up,
             constrain_signs=constrain_signs,
             verbose=verbose,
+            return_fig=return_fig
         )
 
         sf_size.final_size = final_size
@@ -207,6 +214,7 @@ def explain_changes_in_totals(
                 plot_is_static=plot_is_static,
                 width=width,
                 height=height,
+                return_fig=return_fig
             )
         )
         return sp
@@ -233,7 +241,7 @@ def explain_changes_in_totals(
         sf.post_total = df2[total_name].sum()
 
         sf.plot = lambda plot_is_static=False, width=1000, height=1000: plot_waterfall(
-            sf, plot_is_static=plot_is_static, width=width, height=height
+            sf, plot_is_static=plot_is_static, width=width, height=height, return_fig=return_fig
         )
         sf.task = "changes in totals"
         return sf
@@ -252,6 +260,7 @@ def explain_levels(
     verbose=0,
     force_add_up: bool = False,
     constrain_signs: bool = True,
+    return_fig: bool = False
 ):
     """
     Find segments whose average is most different from the global one
@@ -267,6 +276,7 @@ def explain_levels(
     @param verbose: If set to a truish value, lots of debug info is printed to console
     @param force_add_up: Force the contributions of chosen segments to add up to zero
     @param constrain_signs: Whether to constrain weights of segments to have the same
+    @param return_fig: If set to true, plot returns the figure object, otherwise shows the figures
     sign as naive segment averages
     @return: A fitted object
     """
@@ -304,7 +314,7 @@ def explain_levels(
         s["total"] += average * s["seg_size"]
     # print(average)
     sf.reg.intercept_ = average
-    sf.plot = lambda plot_is_static=False, width=2000, height=500, return_fig=False: plot_segments(
+    sf.plot = lambda plot_is_static=False, width=2000, height=500, return_fig=return_fig: plot_segments(
         sf, plot_is_static=plot_is_static, width=width, height=height, return_fig=return_fig
     )
     sf.task = "levels"