From 47b67816e272c9a719b45c0a319321da448b260b Mon Sep 17 00:00:00 2001
From: Andre Wohnsland <50302161+AndreWohnsland@users.noreply.github.com>
Date: Wed, 16 Nov 2022 18:01:03 +0100
Subject: [PATCH] Added option to filter by serving count

---
 frontend/data.py  | 11 +++++++----
 frontend/plots.py |  3 ++-
 frontend/views.py | 13 ++++++++++---
 3 files changed, 19 insertions(+), 8 deletions(-)

diff --git a/frontend/data.py b/frontend/data.py
index 5fcf5ca..af27b3e 100644
--- a/frontend/data.py
+++ b/frontend/data.py
@@ -1,6 +1,5 @@
 import os
 import datetime
-from dataclasses import dataclass
 import streamlit as st
 from deta import Deta
 import pandas as pd
@@ -37,6 +36,7 @@ class ReceivedData():
 
 
 def __myround(x, base=5):
+    """Rounds to the nearest number to given base"""
     return base * round(x / base)
 
 
@@ -152,7 +152,7 @@ def time_aggregation(df: pd.DataFrame, hour_grouping: bool, machine_grouping: bo
 
 
 @st.cache(ttl=60)
-def serving_aggreation(df: pd.DataFrame, machine_split: bool):
+def serving_aggreation(df: pd.DataFrame, machine_split: bool, min_count: int):
     """Aggregates by serving sizes"""
     # rounds to the closest 25
     serving_df = df.copy(deep=True)
@@ -164,7 +164,10 @@ def serving_aggreation(df: pd.DataFrame, machine_split: bool):
         serving_df.groupby(grouping)[DataSchema.language]  # type: ignore
         .agg(["count"])
         .reset_index()
-        .sort_values(['count'], ascending=False)
+        .sort_values([DataSchema.volume], ascending=True)
         .rename(columns={"count": DataSchema.cocktail_count, })
     )
-    return serving_df
+    # for multiple grouping needs to calculate the sum per group and only inlcude the ones having more than min
+    serving_sive_count = serving_df.groupby(DataSchema.volume).sum()
+    volumes_to_keep = serving_sive_count[serving_sive_count[DataSchema.cocktail_count] >= min_count].index.to_list()
+    return serving_df[serving_df[DataSchema.volume].isin(volumes_to_keep)]
diff --git a/frontend/plots.py b/frontend/plots.py
index fca330a..9295548 100644
--- a/frontend/plots.py
+++ b/frontend/plots.py
@@ -147,6 +147,7 @@ def generate_serving_size_bars(df: pd.DataFrame, machine_split: bool):
     fig.update_xaxes(
         tickmode='array',
         tickvals=df[DataSchema.volume],
-        ticktext=[f"{x} ml" for x in df[DataSchema.volume].to_list()]
+        ticktext=[f"{x} ml" for x in df[DataSchema.volume].to_list()],
+        type="category",
     )
     st.plotly_chart(fig, use_container_width=True, config={"displayModeBar": False})
diff --git a/frontend/views.py b/frontend/views.py
index 594e425..756b56a 100644
--- a/frontend/views.py
+++ b/frontend/views.py
@@ -84,7 +84,7 @@ def display_introduction(df_stats: DataFrameStats):
         __what_is_this()
     st.markdown(
         f"""
-        # 📈 Current CocktailBerry Stats
+        ## 📈 Current CocktailBerry Stats
         - 🍸 **{df_stats.cocktails}** cocktails made
         - 🧾 **{df_stats.recipes}** different recipes tasted
         - 🎊 **{df_stats.volume:.1f}** litre cocktails produced
@@ -157,8 +157,15 @@ def __show_volume_stats(filterd_df: pd.DataFrame):
 def __show_serving_size(filterd_df: pd.DataFrame):
     """Show stats over the prepared volume choices"""
     st.header("🥃 Serving Sizes")
-    machine_split = st.checkbox("Split by Machine", False, key="serving_machine")
-    serving_df = data.serving_aggreation(filterd_df, machine_split)
+    col1, col2 = st.columns(2)
+    machine_split = col1.checkbox("Split by Machine", False, key="serving_machine")
+    # only make it available if no machine split is activated
+    max_value_posssible = 10
+    min_servings: int = col2.slider(
+        "Filter Minimal Serving Count", 0,
+        max_value_posssible
+    )  # type: ignore
+    serving_df = data.serving_aggreation(filterd_df, machine_split, min_servings)
     plots.generate_serving_size_bars(serving_df, machine_split)