Payment Analysis

Number of Payments

pb.configure(
    df = df_payments
    , metric = 'payment_sequential'
    , metric_label = 'Share of Payments'
    , metric_label_for_distribution = 'Number of Payments' 
    , agg_func = 'count'
    , norm_by = 'all'
    , axis_sort_order='descending'
    , text_auto='.1%'
    , update_fig={'xaxis': {'tickformat': '.0%'}}
)

print(f'Total number of payments in order: {df_payments.payment_sequential.count():,}')

Total number of payments in order: 103,520

Let’s look at the statistics and distribution of the number of payments per day.

tmp_df_res = (
    df_orders.merge(df_payments, on='order_id', how='left')
    .groupby(pd.Grouper(key='order_purchase_dt', freq='D'), observed=False)['payment_sequential']
    .count()
    .to_frame('payments_cnt_per_day')
)

tmp_df_res['payments_cnt_per_day'].explore.info(
    labels=dict(payments_cnt_per_day='Number of Payments per Day')
    , title='Distribution of Number of Payments per Day'
)

Summary Statistics for "payments_cnt_per_day" (Type: Integer)

Summary			Percentiles			Detailed Stats			Value Counts
Total	604 (100%)		Max	1.21k		Mean	171.39		202	10 (2%)
Missing	---		99%	379		Trimmed Mean (10%)	166.52		160	9 (1%)
Distinct	268 (44%)		95%	312.55		Mode	202		80	7 (1%)
Non-Duplicate	102 (17%)		75%	229		Range	1213		147	7 (1%)
Duplicates	336 (56%)		50%	160		IQR	121		138	6 (<1%)
Dup. Values	166 (27%)		25%	108		Std	93.56		204	6 (<1%)
Zeros	---		5%	47.15		MAD	84.51		108	6 (<1%)
Negative	---		1%	8.12		Kurt	24.75		122	5 (<1%)
Memory Usage	<1 Mb		Min	1		Skew	2.58		128	5 (<1%)

Key Observations:

• 75% of days have ≤230 payments

• Top 5% have ≥312 payments

• Several days exceeded 500 payments

Let’s look at top days.

tmp_df_res.sort_values('payments_cnt_per_day', ascending=False).head()

	payments_cnt_per_day
order_purchase_dt
2017-11-24	1214
2017-11-25	538
2017-11-27	420
2017-11-26	411
2017-11-28	393

Key Observations:

• Black Friday had anomalous payment volumes

By Payment Type

pb.bar_groupby(y='payment_type', to_slide=True)

Key Observations:

• Payment method distribution:

  • Credit card: 74%

  • Boleto: 19%

  • Voucher: 5.5%

  • Debit card: 1.5%

Payment Value

pb.configure(
    df = df_payments
    , metric = 'payment_value'
    , metric_label = 'Average Payment Value, R$'
    , metric_label_for_distribution = 'Payment Value, R$'
    , agg_func = 'mean'
    , axis_sort_order='descending'
    , text_auto='.3s'
)

Let’s see at statistics and distribution of the metric.

pb.metric_info(
    upper_quantile=0.95
    , hist_mode='dual_hist_trim'        
)

Summary Statistics for "payment_value" (Type: Float)

Summary			Percentiles			Detailed Stats			Value Counts
Total	103.52k (100%)		Max	13.66k		Mean	154.02		50	323 (<1%)
Missing	---		99%	1.04k		Trimmed Mean (10%)	114.80		20	272 (<1%)
Distinct	28.99k (28%)		95%	437.19		Mode	50		100	254 (<1%)
Non-Duplicate	13.06k (13%)		75%	171.77		Range	13.66k		77.57	250 (<1%)
Duplicates	74.53k (72%)		50%	100		IQR	114.99		35	165 (<1%)
Dup. Values	15.93k (15%)		25%	56.78		Std	217.46		73.34	160 (<1%)
Zeros	8 (<1%)		5%	26.13		MAD	76.40		30	133 (<1%)
Negative	---		1%	6.71		Kurt	242.72		116.94	131 (<1%)
Memory Usage	1		Min	0		Skew	9.27		56.78	122 (<1%)

Key Observations:

• 75% of payments are ≤172 R$

• Top 5% are ≥440 R$

• Several exceed 6,000 R$

By Payment Type

pb.bar_groupby(y='payment_type', to_slide=True)

Key Observations:

• Credit card payments have highest average value

• Vouchers have lowest

Number of Payment Installments

pb.configure(
    df = df_payments
    , metric = 'payment_installments'
    , metric_label = 'Average Payment Installments'
    , metric_label_for_distribution = 'Payment Installments'
    , agg_func = 'mean'
    , axis_sort_order='descending'
    , text_auto='.2s'
)

Let’s see at statistics and distribution of the metric.

pb.metric_info()

Summary Statistics for "payment_installments" (Type: Integer)

Summary			Percentiles			Detailed Stats			Value Counts
Total	103.52k (100%)		Max	24		Mean	2.85		1	52.38k (51%)
Missing	---		99%	10		Trimmed Mean (10%)	2.28		2	12.38k (12%)
Distinct	23 (<1%)		95%	10		Mode	1		3	10.42k (10%)
Non-Duplicate	2 (<1%)		75%	4		Range	23		4	7.07k (7%)
Duplicates	103.50k (99%)		50%	1		IQR	3		10	5.29k (5%)
Dup. Values	21 (<1%)		25%	1		Std	2.69		5	5.22k (5%)
Zeros	---		5%	1		MAD	0		8	4.26k (4%)
Negative	---		1%	1		Kurt	2.56		6	3.90k (4%)
Memory Usage	1		Min	1		Skew	1.66		7	1.61k (2%)

Key Observations:

• 51% of payments are single-installment

• Some have ≥9 installments

By Payment Type

pb.bar_groupby(y='payment_type', to_slide=True)

Key Observations:

• Credit cards average 3.5 installments

• Other methods have no installments