Add task for plots.

Author: MarenBermudezBoeckle

README.md

@@ -34,7 +34,8 @@ years to observe trends and patterns.
 
 The cleaning part was done in a way in which one can easily adapt the code such that one
 can add other variables of the survey to the survey to the different dictionaries in the
-parampeters.py file.
+parampeters.py file, this is why it does not directly follow exactly the same structure
+as in class where we used specific variables.
 
 Therefore some parts of the cleaning where split up in two parts such that the code
 remains flexible. Since you told us in the assignment that it is fine for readability

crime_increase_perception_commune

@@ -55,14 +55,15 @@ def _fill_missing(enusc_mapped):
     replacements = {99: pd.NA, 77: "Other", 88: "Don't know", 85: "Doesn't apply"}
     enusc_filling = enusc_mapped.copy()
     for column in categories:
+        enusc_filling[column] = enusc_filling[column].astype(pd.Int8Dtype())
         enusc_filling[column] = enusc_filling[column].apply(
             lambda x: replacements.get(x, x) if x in replacements else x
         )
 
         if enusc_mapped[column].dtype != "object":
             enusc_filling[column] = (
                 enusc_filling[column]
-                # .astype(pd.Int8Dtype())
+                # .astype(pd.Int8Dtype()) esto se puede borrar
                 .astype(pd.CategoricalDtype())
             )
             enusc_filling[column] = enusc_filling[column].cat.rename_categories(
@@ -94,5 +95,4 @@ def _set_data_types_not_mapped_var(enusc_filled):
         enusc_dtypes[val] = enusc_dtypes[val].astype(pd.CategoricalDtype())
     for ent in strings:
         enusc_dtypes[ent] = enusc_dtypes[ent].astype(str)
-    enusc_dtypes = enusc_dtypes
     return enusc_dtypes

crime_increase_perception_nation

@@ -0,0 +1,48 @@
+import pandas as pd
+import plotly.graph_objects as go
+
+from project_mbb.config import BLD
+
+products = [
+    BLD / "final" / "perception_plot_commune.png",
+    BLD / "final" / "perception_plot_nation.png",
+    BLD / "final" / "perception_plot_neighborhood.png",
+]
+
+
+def plot_crime_perception(enusc_clean):
+    crime_columns = {
+        "crime_increase_perception_nation": "Crime Perception - Nation",
+        "crime_increase_perception_commune": "Crime Perception - Commune",
+        "crime_increase_perception_neighborhood": "Crime Perception - Neighbourhood",
+    }
+
+    category_labels = ["Decreased", "Stayed the same", "Increased", "Don't know"]
+    for (col, title), i in zip(crime_columns.items(), products, strict=False):
+        enusc_model = pd.DataFrame()
+        enusc_model[col] = enusc_clean[col].cat.codes
+        category_counts = (
+            enusc_model[col].value_counts().reindex([0, 3, 2, 1], fill_value=0)
+        )
+
+        fig = go.Figure()
+
+        fig.add_trace(
+            go.Bar(
+                x=category_labels,
+                y=category_counts.values,
+                marker={"color": ["red", "blue", "green", "gray"]},
+                text=category_counts.values,
+                textposition="auto",
+            )
+        )
+
+        fig.update_layout(
+            title=title,
+            xaxis_title="Perception Category",
+            yaxis_title="Count",
+            template="plotly_white",
+        )
+
+        # Show figure
+        fig.write_image(i)

crime_increase_perception_neighborhood

@@ -0,0 +1,17 @@
+import pandas as pd
+
+from project_mbb.config import BLD
+from project_mbb.final.plot import plot_crime_perception
+
+products = [
+    BLD / "final" / "perception_plot_commune.png",
+    BLD / "final" / "perception_plot_nation.png",
+    BLD / "final" / "perception_plot_neighborhood.png",
+]
+
+
+def task_plot_perception(
+    enusc_clean=BLD / "data" / "enusc_clean.pkl", produces=products
+):
+    enusc_clean = pd.read_pickle(enusc_clean)
+    plot_crime_perception(enusc_clean)