/shipping-analysis [file.xlsx] --company "Company Name"

# Detection logic
if 'Label Processed On' in df.columns:
    date_col = 'Label Processed On'  # Priority - usually has data
else:
    # Fall back to keyword search
    for col in df.columns:
        if 'date' in col.lower() or 'processed' in col.lower():
            # Verify column has actual date values
            test = pd.to_datetime(df[col].dropna().head(10), errors='coerce')
            if test.notna().sum() > 5:
                date_col = col
                break

| Metrics (A-B)          | Top Origins (D-F)      | Weight Bands (G-I)     |
|------------------------|------------------------|------------------------|
| Total Shipments: X     | Client: Count (%)      | 1 oz: Count (%)        |
| Date Range: X to Y     | Client: Count (%)      | 2 oz: Count (%)        |
| Days Covered: X        | Client: Count (%)      | ...                    |
| Avg Monthly Volume: X  | ...                    | 15.99 oz: Count (%)    |
| Unique Clients: X      |                        | 1 lb: Count (%)        |
| Avg Weight (lb): X     |                        | 2 lb: Count (%)        |
| Median Weight (lb): X  |                        | ...                    |
| Avg Weight (oz): X     |                        | 50 lb: Count (%)       |
| Median Weight (oz): X  |                        | 50+ lbs: Count (%)     |

1 oz      181     0.35%
2 oz      194     0.37%
3 oz      937     1.79%
...
15 oz      70     0.13%
15.99 oz  288     0.55%
1 lb    1,768     3.39%
2 lb    1,570     3.01%
...
50 lb       6     0.01%
50+ lbs    15     0.03%

weight_data = []
# Individual ounces 1-15
for oz in range(1, 16):
    count = len(df[(df['Weight_oz'] > oz-1) & (df['Weight_oz'] <= oz)])
    if count > 0:
        weight_data.append((f"{oz} oz", count, count/total))

# 15.99 oz (15-16 oz boundary)
count_1599 = len(df[(df['Weight_oz'] > 15) & (df['Weight_oz'] <= 16)])
if count_1599 > 0:
    weight_data.append(("15.99 oz", count_1599, count_1599/total))

# Individual pounds 1-50
for lb in range(1, 51):
    lb_oz_low = lb * 16
    lb_oz_high = (lb + 1) * 16
    count = len(df[(df['Weight_oz'] > lb_oz_low) & (df['Weight_oz'] <= lb_oz_high)])
    if count > 0:
        weight_data.append((f"{lb} lb", count, count/total))

# 50+ lbs
count_50plus = len(df[df['Weight_oz'] > 800])
if count_50plus > 0:
    weight_data.append(("50+ lbs", count_50plus, count_50plus/total))

def format_pct(val):
    """Format decimal as percentage: 0.792 -> '79.22%'"""
    if isinstance(val, (int, float)):
        return f"{val * 100:.2f}%"
    return val

def format_num(val):
    """Format number with commas: 41368 -> '41,368'"""
    if isinstance(val, (int, float)):
        return f"{int(val):,}"
    return val

from openpyxl.cell.cell import MergedCell

for row in ws.iter_rows():
    for cell in row:
        if isinstance(cell, MergedCell):
            continue  # Skip - MergedCells don't have normal properties
        cell.alignment = center_align
        cell.border = thin_border

# Drop temp columns before Raw Data Sample sheet
temp_cols = ['_Month', 'Weight_oz', 'Weight_lb', 'DIM_Weight']
sample = df.drop(columns=[c for c in temp_cols if c in df.columns], errors='ignore')

File: PLD 9.12 - 10.12.xlsx (+ 2 more PLD files)
Company: F.R.O.G. Fulfillment

Column Mapping:
- Date: "Label Processed On" (NOT "Ship Date" - it's empty)
- Origin Proxy: "Client Name" (they're a 3PL)
- Service Level: "Service Code Description"
- Carrier: "Carrier"
- Weight (lbs): "Weight"
- Destination State: "Ship To State"
- Dimensions: "Box Length", "Box Width", "Box Height"

Output: 11-sheet workbook with 3-column Executive Summary

File: Client_Export.xlsx
Company: Beta Brands

Column Mapping:
- Order Date: Column B
- Ship Address: Column G
- Service Level: Column K
- Weight (oz): Column R
- DIMS: L(S), W(T), H(U)

No cost or delivery data. Run core + DIM exposure.