Grades
Grades of the weekly advances, on USM's 0–100 scale (55 is a pass). Press Reveal on your team to see its grade and point-by-point feedback against the checklist on the weekly progress page, with advice for the next advance. If your team is missing or something looks wrong, write to hizocar@gmail.com.
Advance 1 · Team, data, and question
Due Tuesday, November 3, 23:59. The checklist has five items: team registered, notebook shared, data loaded from its URL, first inspection, and business question with its user.
Casa Central
Cuando pagan
85 🟢
- ✅Registered in the mailbox. Team and case registered.
- ✅Notebook in Colab, shared. Shared with the instructor; the first cell names the team, members, and case.
- ✅Data loaded from its URL.
employee_compensation.csvread from the course URL. - ✅First inspection. Shape, types, first rows, missing values (8 of 9 columns), and 97 duplicates, plus
describe()and the list of labels in every text column. Your “Qué observamos” cell already spots the inconsistent labels (PhD/Phd/PHD, FINANCE/Finance, extra spaces) and the impossible values (Age 119, Years_Experience 78, Performance_Score 99). - ✅Business question and its user. Specific and useful: what determines salary, to detect gaps without objective justification, for HR and the compensation team, with
Salaryas the target.
How to improve
- The work deserves 100; the grade is 85 because it arrived one day late (−15). Mind Tuesday's deadline for Advance 2.
- Your list of problems is the starting point of Advance 2: turn each one into a line of the cleaning log (problem · decision · rows affected).
Los Alemanes
93 🟢
- ✅Registered in the mailbox. Team and case registered.
- ⚠️Notebook in Colab, shared. Shared, but the notebook is called Untitled1.ipynb. Give it your team's name, and delete the empty cell at the end.
- ✅Data loaded from its URL. Read from the course URL through a
DATA_URLvariable: good practice. - ✅First inspection. Shape, types, missing values, duplicates, and first rows in one tidy cell, with a written summary (10,100 rows, 97 duplicates, 8 of 9 columns with gaps, Performance_Score the worst with 603).
- ⚠️Business question and its user. It is clear and names its user, but it repeats the brief almost word for word. Make it yours: which decision will the people team take with the answer, and which gaps matter most to them (department, region, gender?).
How to improve
- Add to your summary what the inspection already shows: inconsistent labels (
bachelorvsBachelor,FINANCE," West") and impossible values (Age up to 119, Years_Experience −10 to 78, Performance_Score up to 99). Those are the problems you fix in Advance 2.
búsqueda inusual
93 🟢
- ✅Registered in the mailbox. Team and case registered.
- ✅Notebook in Colab, shared. Shared, with the team and members at the top.
- ✅Data loaded from its URL.
employee_compensation.csvread from the course URL. - ⚠️First inspection. All five checks are there (shape, types, missing values, 97 duplicates, first rows), but there is no written reading of what they show.
- ✅Business question and its user. Clear and focused: the effect of experience, education, and performance on salary, for TalentCo's HR management to design competitive pay and retain talent.
How to improve
- Add a short “What we see” cell after the inspection: 97 duplicate rows, missing values in 8 of 9 columns, and labels written in several ways (
bachelor/Bachelor,FINANCE/Finance). Numbers plus one sentence on why each matters. - Run
df.describe()too: it reveals impossible values (Age 119, Performance_Score 99) that you will need to fix in Advance 2.
Ana y Cléo
90 🟢
- ✅Registered in the mailbox. Team and case registered.
- ✅Notebook in Colab, shared. Shared with the instructor.
- ✅Data loaded from its URL.
employee_compensation.csvread from the course URL. - ⚠️First inspection. Every check is there (shape, types,
info(), first rows, missing values, 97 duplicates), but the cells were run out of order: the cells that import pandas and load the data have no run recorded. Use Runtime › Restart session and run all before submitting. The text cell copies the checklist instead of saying what you found. - ✅Business question and its user. Good angle: which factors drive salary and whether there is discrimination in pay, for HR to detect and fix it.
How to improve
- Replace the copied checklist line with your own observations: how many duplicates, which columns have gaps, which labels are inconsistent.
- Write the notebook in one language (it mixes English and Spanish); either is fine.
Grupito
63 🟡
- ⚠️Registered in the mailbox. Registered, but the advance was submitted in the mailbox as Advance 2. Choose the right advance when you submit.
- ✅Notebook in Colab, shared. Shared with the instructor.
- ❌Data loaded from its URL. The notebook loads
cordillera_mobile.csv, the dataset of the classes, instead of your case'semployee_compensation.csv. - ⚠️First inspection. The five checks are done (5,075 rows, 75 duplicates…), but they describe the wrong data, so none of the findings apply to TalentCo.
- ✅Business question and its user. Well written: the main factors behind salaries and how much each explains, for HR and TalentCo's management to review compensation policies.
How to improve
- Change the URL to
https://raw.githubusercontent.com/hizocar/usm-course-intensivo/main/docs/datasets/employee_compensation.csvand run everything again before starting Advance 2: the cleaning has to be done on your case's data. - Add one sentence per check saying what it shows.
Ignachelo
Advance 1 pending: write to hizocar@gmail.com to agree how your team catches up.
Vitacura
FinTrust Bank Vicente Célia
100 🟢
- ✅Registered in the mailbox. Team and case registered (as “Vincente”: use one spelling of the team name everywhere).
- ✅Notebook in Colab, shared. Shared, with a clear title, the case, and the members.
- ✅Data loaded from its URL.
loan_default.csvread from the course URL. - ✅First inspection. Shape (10,060 × 16), types, first rows, missing values per column, and 60 duplicates, each followed by a written analysis. You went further:
interest_rateis stored as text and 803 rows (7.98%) carry a “%” sign. - ✅Business question and its user. In your own words, with its users (the risk department and the credit analysts) and
defaultas the target.
How to improve
- For Advance 2, also look for inconsistent labels (
full_timevsFull Time,Auto/car) and impossible values (credit scores outside 300–850). - Before submitting, run Restart session and run all: several cells have no run recorded.
Mafilia
0 ⚫
- ❌Advance 1. Nothing was received in the mailbox, and the three-day late window closed on Friday, November 6.
How to improve
- Advance 2 is due Tuesday, November 10, and it builds on Advance 1: load your case's data from its URL and run the first inspection before you start cleaning. If something is blocking your team, write to hizocar@gmail.com today.
PJMC
0 ⚫
- ❌Advance 1. Nothing was received in the mailbox, and the three-day late window closed on Friday, November 6.
How to improve
- Advance 2 is due Tuesday, November 10, and it builds on Advance 1: load your case's data from its URL and run the first inspection before you start cleaning. If something is blocking your team, write to hizocar@gmail.com today.
The Boys
0 ⚫
- ❌Advance 1. Nothing was received in the mailbox, and the three-day late window closed on Friday, November 6.
How to improve
- Advance 2 is due Tuesday, November 10, and it builds on Advance 1: load your case's data from its URL and run the first inspection before you start cleaning. If something is blocking your team, write to hizocar@gmail.com today.
| Grade | Meaning | |
|---|---|---|
| 85 – 100 | 🟢 | Every item is done and correct. |
| 55 – 84 | 🟡 | Every item is there, but some need fixing before the next advance. |
| 1 – 54 | 🔴 | Important items are missing. |
| 0 | ⚫ | Not submitted. |
Late submissions lose 15 points per day for up to three days; after that the advance gets a 0.