|
| 1 | +# Licensed to the Apache Software Foundation (ASF) under one |
| 2 | +# or more contributor license agreements. See the NOTICE file |
| 3 | +# distributed with this work for additional information |
| 4 | +# regarding copyright ownership. The ASF licenses this file |
| 5 | +# to you under the Apache License, Version 2.0 (the |
| 6 | +# "License"); you may not use this file except in compliance |
| 7 | +# with the License. You may obtain a copy of the License at |
| 8 | +# |
| 9 | +# http://www.apache.org/licenses/LICENSE-2.0 |
| 10 | +# |
| 11 | +# Unless required by applicable law or agreed to in writing, |
| 12 | +# software distributed under the License is distributed on an |
| 13 | +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY |
| 14 | +# KIND, either express or implied. See the License for the |
| 15 | +# specific language governing permissions and limitations |
| 16 | +# under the License. |
| 17 | + |
| 18 | +""" |
| 19 | +This query counts how many customers within a specific range of country codes have not placed |
| 20 | +orders for 7 years but who have a greater than average “positive” account balance. It also reflects |
| 21 | +the magnitude of that balance. Country code is defined as the first two characters of c_phone. |
| 22 | +""" |
| 23 | + |
| 24 | +from datafusion import SessionContext, WindowFrame, col, lit, functions as F |
| 25 | + |
| 26 | +NATION_CODE = 13 |
| 27 | + |
| 28 | +# Load the dataframes we need |
| 29 | + |
| 30 | +ctx = SessionContext() |
| 31 | + |
| 32 | +df_customer = ctx.read_parquet("data/customer.parquet").select_columns( |
| 33 | + "c_phone", "c_acctbal", "c_custkey" |
| 34 | +) |
| 35 | +df_orders = ctx.read_parquet("data/orders.parquet").select_columns("o_custkey") |
| 36 | + |
| 37 | +# The nation code is a two digit number, but we need to convert it to a string literal |
| 38 | +nation_code = lit(str(NATION_CODE)) |
| 39 | + |
| 40 | +# Use the substring operation to extract the first two charaters of the phone number |
| 41 | +df = df_customer.with_column("cntrycode", F.substr(col("c_phone"), lit(0), lit(3))) |
| 42 | + |
| 43 | +# Limit our search to customers with some balance and in the country code above |
| 44 | +df = df.filter(col("c_acctbal") > lit(0.0)) |
| 45 | +df = df.filter(nation_code == col("cntrycode")) |
| 46 | + |
| 47 | +# Compute the average balance. By default, the window frame is from unbounded preceeding to the |
| 48 | +# current row. We want our frame to cover the entire data frame. |
| 49 | +window_frame = WindowFrame("rows", None, None) |
| 50 | +df = df.with_column( |
| 51 | + "avg_balance", F.window("avg", [col("c_acctbal")], window_frame=window_frame) |
| 52 | +) |
| 53 | + |
| 54 | +# Limit results to customers with above average balance |
| 55 | +df = df.filter(col("c_acctbal") > col("avg_balance")) |
| 56 | + |
| 57 | +# Limit results to customers with no orders |
| 58 | +df = df.join(df_orders, (["c_custkey"], ["o_custkey"]), "anti") |
| 59 | + |
| 60 | +# Count up the customers and the balances |
| 61 | +df = df.aggregate( |
| 62 | + [col("cntrycode")], |
| 63 | + [ |
| 64 | + F.count(col("c_custkey")).alias("numcust"), |
| 65 | + F.sum(col("c_acctbal")).alias("totacctbal"), |
| 66 | + ], |
| 67 | +) |
| 68 | + |
| 69 | +df = df.sort(col("cntrycode").sort()) |
| 70 | + |
| 71 | +df.show() |
0 commit comments