Create app.py

app.py

@@ -0,0 +1,154 @@
+import gradio as gr
+import os
+import pandas as pd
+from fpdf import FPDF
+import matplotlib.pyplot as plt
+import seaborn as sns
+import numpy as np
+
+# Add your PDFReport class and generate_data_report function here
+class PDFReport(FPDF):
+    def header(self):
+        self.set_font('Arial', 'B', 12)
+        self.cell(0, 10, 'Data Exploration Report', border=1, ln=1, align='C')
+        self.ln(10)
+
+    def chapter_title(self, title):
+        self.set_font('Arial', 'B', 12)
+        self.cell(0, 10, title, border=1, ln=1, align='C')
+        self.ln(5)
+
+    def chapter_body(self, text):
+        self.set_font('Arial', '', 10)
+        self.multi_cell(0, 10, text)
+        self.ln()
+
+    def add_table(self, headers, data, col_widths):
+        self.set_font('Arial', 'B', 10)
+        for idx, header in enumerate(headers):
+            self.cell(col_widths[idx], 10, header, border=1, align='C')
+        self.ln()
+        self.set_font('Arial', '', 10)
+        for row in data:
+            for idx, item in enumerate(row):
+                self.cell(col_widths[idx], 10, str(item), border=1)
+            self.ln()
+def generate_data_report(data,output_file='data_report.pdf', selected_columns=None):
+    if isinstance(data, str):
+        file_path = data
+        file_extension = os.path.splitext(file_path)[1].lower()
+        if file_extension == '.csv':
+            file_format = 'CSV'
+        elif file_extension in ['.xls', '.xlsx']:
+            file_format = 'Excel'
+        else:
+            file_format = 'Unknown format'
+        if file_format == 'CSV':
+            data = pd.read_csv(file_path)
+        elif file_format == 'Excel':
+            data = pd.read_excel(file_path)
+    else:
+        file_format = 'DataFrame'
+        file_path = "DataFrame"
+
+    pdf = PDFReport()
+    pdf.add_page()
+
+    pdf.set_font('Arial', 'B', 12)
+    pdf.cell(0, 10, f"File Name: {os.path.basename(file_path)}", ln=True, align='L')
+    pdf.cell(0, 10, f"File Format: {file_format}", ln=True, align='L')
+    pdf.cell(0, 10, f"Total Data: {data.shape[0]} rows, {data.shape[1]} columns", ln=True, align='L')
+    pdf.ln(10)
+
+    pdf.chapter_title("Columns with Missing Values")
+    total_values = len(data)
+    missing_values = data.isnull().sum()
+    missing_cols = [
+        [col, total_values, missing_values[col]]
+        for col in missing_values[missing_values > 0].index
+    ]
+    if missing_cols:
+        pdf.add_table(["Column Name", "Total Values", "Missing Values"], missing_cols, [100, 40, 50])
+    else:
+        pdf.chapter_body("No columns with missing values.")
+
+    pdf.chapter_title("Columns Categorized by Data Type")
+    dtypes_summary = data.dtypes.value_counts().reset_index()
+    dtypes_summary.columns = ['Data Type', 'Count']
+    pdf.add_table(["Data Type", "Count"], dtypes_summary.values.tolist(), [100, 50])
+
+    column_types = {}
+    for dtype in data.dtypes.unique():
+        column_types[str(dtype)] = data.select_dtypes(include=[dtype]).columns.tolist()
+
+    for dtype, columns in column_types.items():
+        pdf.chapter_title(f"Columns of Type: {dtype}")
+        col_data = [[col] for col in columns]
+        pdf.add_table(["Column Name"], col_data, [190])
+
+    pdf.chapter_title("Constant Columns")
+    constant_cols = [col for col in data.columns if data[col].nunique() == 1]
+    if constant_cols:
+        constant_cols_data = [[col] for col in constant_cols]
+        pdf.add_table(["Constant Column Name"], constant_cols_data, [190])
+        data = data.drop(columns=constant_cols)
+        pdf.chapter_body("Constant Columns After Removal: None")
+    else:
+        pdf.chapter_body("No constant columns found.")
+
+    pdf.chapter_title("Box Plots for Numeric Columns")
+    numeric_cols = data.select_dtypes(include=np.number).columns
+    boxplot_dir = "box_plots"
+    os.makedirs(boxplot_dir, exist_ok=True)
+
+    boxplot_colors = ['#FF6347', '#3CB371', '#8A2BE2', '#FF4500', '#1E90FF', '#FFD700']
+
+    for idx, col in enumerate(numeric_cols):
+        plt.figure(figsize=(6, 4))
+        sns.boxplot(x=data[col], color=boxplot_colors[idx % len(boxplot_colors)])
+        plt.title(f"Box Plot: {col}")
+        plt.savefig(f"{boxplot_dir}/{col}.png")
+        plt.close()
+        pdf.add_page()
+        pdf.chapter_title(f"Box Plot: {col}")
+        pdf.image(f"{boxplot_dir}/{col}.png", w=170)
+
+    pdf.chapter_title("Distribution Charts")
+    dist_dir = "distribution_charts"
+    os.makedirs(dist_dir, exist_ok=True)
+    if selected_columns is None:
+        selected_columns = data.columns[:6]
+
+    dist_colors = ['#8B0000', '#228B22', '#DAA520', '#B0C4DE', '#9932CC', '#FF69B4']
+
+    for idx, col in enumerate(selected_columns):
+        plt.figure(figsize=(6, 4))
+        if col in numeric_cols:
+            sns.histplot(data[col], kde=True, color=dist_colors[idx % len(dist_colors)])
+        else:
+            data[col].value_counts().plot(kind='bar', color=dist_colors[idx % len(dist_colors)])
+        plt.title(f"Distribution of {col}")
+        plt.savefig(f"{dist_dir}/{col}.png")
+        plt.close()
+        pdf.add_page()
+        pdf.chapter_title(f"Distribution: {col}")
+        pdf.image(f"{dist_dir}/{col}.png", w=170)
+
+    pdf.output(output_file)
+    print(f"Report saved as {output_file}")
+def generate_report(file):
+    file_path = file.name
+    output_file = "data_report.pdf"
+    generate_data_report(file_path, output_file=output_file)
+    return output_file
+
+iface = gr.Interface(
+    fn=generate_report,
+    inputs=gr.File(label="Upload Dataset (.csv or .xlsx)"),
+    outputs=gr.File(label="Download PDF Report"),
+    title="Data Exploration Tool",
+    description="Upload your dataset to generate a PDF data exploration report."
+)
+
+if __name__ == "__main__":
+    iface.launch()