Py.Cafe

import plotly.express as px
from dash import Dash, dcc, callback, Input, Output
import numpy as np
import pandas as pd


df = pd.read_csv("x-24-followers.csv")
# dff = df[df['Token'].isin(['Qeqe','Kabal','Pepe','Chad'])]

# Remove unwanted columns
data_new = df.drop(columns=['Avg Growth', 'CAGR'])

# Convert all community size columns to numeric
data_new = data_new.apply(lambda col: pd.to_numeric(col.str.replace(',', '', regex=True), errors='coerce')
    if col.name != 'Token' and col.dtype == 'object' else col)

# Explicitly replace -1 values with np.nan to ensure they are treated as missing values for interpolation
data_new.replace(-1, np.nan, inplace=True)
data_new['Token'] = data_new['Token'].astype(str)
# Interpolate missing values linearly
data_new_interpolated = data_new.set_index('Token').T.interpolate(method='linear').T.reset_index()

# Melt the dataframe to long format for plotting
data_long = data_new_interpolated.melt(id_vars=['Token'], var_name='Date', value_name='Community_Size')

# Convert the 'Community_Size' column to numeric, removing commas if present
data_long['Community_Size'] = data_long['Community_Size'].astype(float)

# Convert 'Date' to datetime for better plotting
data_long['Date'] = pd.to_datetime(data_long['Date'])
data_long = data_long.sort_values(by=['Token', 'Date'])

# Calculate the daily percentage growth for each token
data_long['Daily_Growth_Percentage'] = data_long.groupby('Token')['Community_Size'].pct_change(fill_method=None) * 100


app = Dash()
app.layout = [
    dcc.Dropdown(id='my-token',
                 options=[{'label':'all', 'value':'all'}]+[{'label':x, 'value':x} for x in sorted(data_new['Token'])],
                 value=['all'],
                 multi=True,
                 clearable=False),
    dcc.Graph(id='line-graph'),
    dcc.Markdown('Minimum age of community (days)'),
    dcc.Input(id='my-days', type='number', value=9, min=1, max=50, step=1),
    dcc.Graph(id='bar-graph'),
]


@callback(
    Output('line-graph', 'figure'),
    Input('my-token', 'value')
)
def update_graph(tokens):
    if tokens == ['all']:
        fig = px.line(data_long, x='Date', y='Daily_Growth_Percentage',
                      color='Token',
                      title="Daily Community Growth Percentage Over Time by Token",
                      height=750,
                      labels={
                          'Daily_Growth_Percentage': 'Daily Growth (%)',
                          'Date': 'Date'})
        fig.update_traces(mode='lines+markers')
    else:
        data_long_filtered = data_long[data_long['Token'].isin(tokens)]
        fig = px.line(data_long_filtered, x='Date', y='Daily_Growth_Percentage',
                      color='Token',
                      title="Daily Community Growth Percentage Over Time by Token",
                      height=750,
                      labels={
                          'Daily_Growth_Percentage': 'Daily Growth (%)',
                          'Date': 'Date'})
        fig.update_traces(mode='lines+markers')
    return fig

@callback(
    Output('bar-graph', 'figure'),
    Input('my-days', 'value')
)
def update_graph(min_days):
    # Count non-NaN values for Community_Size grouped by Token
    valid_counts = data_long.groupby('Token')['Community_Size'].apply(
        lambda x: x.notna().sum())
    # Filter out Tokens with fewer than x-days valid Community_Size entries
    tokens_to_keep = valid_counts[valid_counts >= min_days].index
    filtered_df = data_long[data_long['Token'].isin(tokens_to_keep)]

    median_growth = filtered_df.groupby("Token")["Daily_Growth_Percentage"].median().reset_index()
    median_growth.rename(columns={"Daily_Growth_Percentage": "median_prcnt_growth"}, inplace=True)

    fig = px.bar(median_growth, x='Token', y='median_prcnt_growth', title='Median growth')
    return fig