Excel BI - PowerQuery Challenge 206

excel-challenges
power-query
Group1 Group2
Published

March 24, 2026

Illustration for Excel BI - PowerQuery Challenge 206

Challenge Description

Group1 Group2

Solutions

library(tidyverse)
library(readxl)

path = "Power Query/PQ_Challenge_206.xlsx"
input = read_excel(path, range = "A1:D13")
test  = read_excel(path, range = "F1:K19")

r1 = input %>%
  mutate(group = cumsum(is.na(Group1)) + 1) %>%
  filter(!is.na(Group1)) %>%
  mutate(nr = row_number(), .by = group) %>%
  unite("Group", Group1:Group2, sep = "-") %>%
  unite("Value", Value1:Value2, sep = "-") %>%
  pivot_longer(-c(nr, group), names_to = "Variable", values_to = "Value") %>%
  select(-Variable)

rearrange_df <- function(df, part) {
  df %>%
    filter(group == part) %>%
    select(-group) %>%
    mutate(col = nr, row = row_number()) %>%
    pivot_wider(names_from = col, values_from = Value) %>%
    as.data.frame()
}

result = map_df(unique(r1$group), ~ rearrange_df(r1, .x)) %>%
  select(-c(1,2)) %>%
  separate_wider_delim(1:ncol(.), delim = "-", names_sep = "-") %>%
  mutate(across(everything(), ~ if_else(. == "NA", NA_character_, .)))

names(result) = names(test)

all.equal(result, test)
# [1] TRUE

  • Logic:

    • Reads the workbook range needed for the challenge

    • Reshapes the data into the structure required by the result table

    • Builds helper columns that drive the final output

  • Strengths:

    • The R solution stays close to the workbook logic and keeps the transformation compact.

  • Areas for Improvement:

    • The code assumes the workbook layout and selected ranges remain stable.

  • Gem:

    • The best part of the solution is choosing the right intermediate shape before formatting the final output.
import pandas as pd
import numpy as np

path = "PQ_Challenge_206.xlsx"
input = pd.read_excel(path, usecols="A:D", nrows = 12)
test = pd.read_excel(path, usecols="F:K", nrows=18).astype(str).replace("nan", np.NaN)

r1 = input.copy()
r1["group"] = r1["Group1"].isna().cumsum() + 1
r1 = r1[~r1["Group1"].isna()]
r1["nr"] = r1.groupby("group").cumcount() + 1
r1["Group"] = r1["Group1"].astype(str) + "-" + r1["Group2"].astype(str)
r1["Value"] = r1["Value1"].astype(str) + "-" + r1["Value2"].astype(str)
r1 = r1.drop(columns=["Group1", "Group2", "Value1", "Value2"])\
    .melt(id_vars=["nr", "group"], value_vars=["Group", "Value"], var_name="Variable", value_name="MeltedValue")\
    .sort_values(by=["group", "nr"]).reset_index(drop=True).drop(columns=["Variable"])

def rearrange_df(df, part):
    df_part = df[df["group"] == part].drop(columns=["group"]).reset_index(drop=True)
    df_part["col"] = df_part["nr"]
    df_part["row"] = df_part.index + 1
    df_part = df_part.pivot(index="row", columns="col", values="MeltedValue").rename(columns=lambda x: "c" + str(x))
    df_part = df_part.map(lambda x: x.split("-", maxsplit=1) if isinstance(x, str) else x)
    return df_part

r2 = pd.concat([rearrange_df(r1, i) for i in r1["group"].unique()], axis=0).reset_index(drop=True)
r2 = pd.concat([r2[col].apply(pd.Series) for col in r2.columns], axis=1)
r2.columns = test.columns
r2 = r2.replace({"nan": np.NaN, "\\.0": ""}, regex=True)

print(r2.equals(test)) # True

  • Logic:

    • Reads the workbook range needed for the challenge

    • Reshapes the data into the structure required by the result table

    • Aggregates or ranks values at the relevant grouping level

    • Applies the rule iteratively until the output is complete

  • Strengths:

    • The Python version follows the same workbook rule in a direct pandas-oriented implementation.

  • Areas for Improvement:

    • As with the R version, any workbook layout change would require small adjustments.

  • Gem:

    • The implementation stays close to the source challenge instead of adding unnecessary abstraction.

Difficulty Level

This task is moderate:

  • It combines reshaping, grouping, or parsing steps that are common in Power Query style problems.

  • The main challenge is reproducing the workbook output structure exactly.