Add plot for Poisson distribution

README.md

@@ -29,7 +29,7 @@ This is the repository for a version of the Distribution Zoo, built in Streamlit
 | Binomial                  |       |          |       |      |      |       |             |        |        |   |      |
 | Discrete-Uniform          |       |          |       |      |      |       |             |        |        |   |      |
 | Negative-Binomial         |       |          |       |      |      |       |             |        |        |   |      |
-| Poisson                   |       |          |       |      |      |       |             |        |        |   |      |
+| Poisson                   | :+1:  |          |       |      |      |       |             |        |        |   |      |
 | **Multivariate**          |       |          |       |      |      |       |             |        |        |   |      |
 | Dirichlet                 |       |          |       |      |      |       |             |        |        |   |      |
 | Inverse-Wishart           |       |          |       |      |      |       |             |        |        |   |      |

distribution_zoo/__init__.py

@@ -12,12 +12,17 @@
     Gamma,
 )
 
+from .disc_uni import (
+    Poisson,
+)
+
 dist_mapping = {
     DistributionClass('Continuous Univariate', 'cont_uni'): [
         Normal(),
         Gamma(),
     ],
     DistributionClass('Discrete Univariate', 'disc_uni'): [
+        Poisson()
     ],
     DistributionClass('Multivariate', 'mult'): [
     ],

distribution_zoo/disc_uni/__init__.py

@@ -0,0 +1 @@
+from .poisson import Poisson

distribution_zoo/disc_uni/poisson.py

@@ -0,0 +1,86 @@
+from distribution_zoo import BaseDistribution
+
+import plotly.graph_objects as go
+import numpy as np
+import pandas as pd
+import scipy.stats as stats
+import streamlit as st
+
+
+class Poisson(BaseDistribution):
+
+    display_name = 'Poisson'
+    range_min = 0
+    range_max = 100
+    param_rate = st.session_state['poisson_rate'] if 'poisson_rate' in st.session_state else 10.0
+    param_range_start = None
+    param_range_end = None
+
+    def __init__(self):
+        super().__init__()
+
+    def sliders(self):
+
+        if 'poisson_range' not in st.session_state:
+            self.update_range()
+
+        # This slider's initial value is set from st.session_state['normal_range'], set with update_range()
+        self.param_range_start, self.param_range_end = st.sidebar.slider(
+            'Range', min_value=self.range_min, max_value=self.range_max, step=1, key='poisson_range'
+        )
+
+        self.param_rate = st.sidebar.slider(
+            r'Rate', min_value=0.0, max_value=32.0, value=self.param_rate, step=0.1, key='poisson_rate',
+            on_change=self.update_range
+        )
+
+    def update_range(self):
+
+        rate = st.session_state['poisson_rate'] if 'poisson_rate' in st.session_state else self.param_rate
+
+        new_lower = int(0)
+        new_upper = int(min(1 + round(stats.poisson(mu=rate).ppf(0.999), 0), self.range_max))
+        st.session_state['poisson_range'] = (new_lower, new_upper)
+
+    def plot(self):
+        x = range(self.param_range_start, self.param_range_end)
+
+        chart_data = pd.DataFrame(
+            {
+                'x': x,
+                'pmf': stats.poisson.pmf(x, mu=self.param_rate),
+                'cdf': stats.poisson.cdf(x, mu=self.param_rate),
+            }
+        )
+
+        line_data = pd.DataFrame(
+            {
+                'x': [self.param_rate, self.param_rate],
+                'pmf': [0.0, max(chart_data['pmf'])],
+                'cdf': [0.0, max(chart_data['cdf'])],
+            }
+        )
+
+        # Create Plotly chart for the PDF
+        pdf_chart = go.Figure(go.Bar(x=chart_data['x'], y=chart_data['pmf'], name='PMF'))
+        pdf_chart.add_trace(go.Scatter(x=line_data['x'], y=line_data['pmf'], mode='lines', name=f'Mean ({self.param_rate})', line=dict(color='orange', width=2)))
+        pdf_chart.update_layout(xaxis_title='x', yaxis_title='pmf', margin=dict(l=20, r=20, t=20, b=20))
+
+        # Create Plotly chart for the CDF
+        cdf_chart = go.Figure(go.Bar(x=chart_data['x'], y=chart_data['cdf'], name='CDF'))
+        cdf_chart.add_trace(go.Scatter(x=line_data['x'], y=line_data['cdf'], mode='lines', name=f'Mean ({self.param_rate})', line=dict(color='orange', width=2)))
+        cdf_chart.update_layout(xaxis_title='x', yaxis_title='cdf', margin=dict(l=20, r=20, t=20, b=20))
+
+        # Streamlit columns for displaying the charts
+        pdf_col, cdf_col = st.columns(2)
+
+        with pdf_col:
+            st.subheader('Probability mass function')
+            st.plotly_chart(pdf_chart, use_container_width=True)
+
+        with cdf_col:
+            st.subheader('Cumulative distribution function')
+            st.plotly_chart(cdf_chart, use_container_width=True)
+
+    def update_code_substitutions(self):
+        pass