Download Cleaned DepEd School Data, SY 2017-18 to 2025-26

This page offers a free download of the cleaned, harmonized school-level datasets we built from the Department of Education’s (DepEd) public files. They hold nine years of enrollment (SY 2017-18 to SY 2025-26) and the SY 2023-24 school inputs. They also hold the National Achievement Test (NAT), Early Language, Literacy, and Numeracy Assessment (ELLNA), and Comprehensive Rapid Literacy Assessment (CRLA) results. The files come in CSV and Parquet with a codebook, checksums, and terms of use that say how you may use them. Version 1.0 is 66.6 MB in total, and there is no sign-up. If you have read our data reports and wanted the files behind them, this page is for you.

The raw DepEd files are public, but they are not ready to analyze. Labels change between years, blanks mean three different things, and one year is encoded in two character sets. The modules also have to be joined on a school ID that a few hundred schools change. The data quality audit lists every issue in the raw files. These files are what you get after every documented rule is applied, so you can start from the same tables the reports use instead of repeating a week of cleaning.

What you get. Five cleaned datasets (343 documented columns), two helper tables, a codebook, the harmonization rules, SHA-256 checksums and a README. The data files follow DepEd’s terms of use; our code and documentation need our permission to reuse (terms of use). The original DepEd files were never edited; every change is reproducible from the published pipeline.

The files in the release

Each dataset comes as a gzip-compressed CSV, which opens in any tool, and as Parquet, which is smaller, typed, and faster in Python and R. The nine-year enrollment panel is one Parquet file or one CSV per school year.

Documentation in the same folder:

Every file is at https://data.nash.com.ph/deped-data/v1/. The folder name is the version, so the contents behind a URL never change. A future release will be a new folder, and this page will list what changed.

Quick start

Excel or Google Sheets. Download a .csv.gz file and unzip it. On Windows, use 7-Zip or tar -xzf in a terminal; on macOS, double-click the file. The school master has 60,167 rows and 207 columns, and each enrollment year has about 60,000 rows, so each fits in a sheet. Blank cells are missing values, not zeros. Keep school_id as text if your tool offers the choice, and do not let it drop leading digits.

Python. Parquet needs pyarrow; CSV needs nothing beyond pandas.

import pandas as pd

base = "https://data.nash.com.ph/deped-data/v1"
master = pd.read_parquet(f"{base}/school_master_2023.parquet")
panel = pd.read_parquet(f"{base}/enrollment_panel_long.parquet")

# Public elementary schools with a pupil-teacher ratio above 40, by region
high = master[master.is_public & master.offers_es & (master.ptr > 40)]
print(high.groupby("region_const").size().sort_values(ascending=False))

# National enrollment by year (schools that reported nothing are excluded by design)
print(panel.groupby("sy_label").enrol_total.sum())

R.

library(arrow)
library(dplyr)

base <- "https://data.nash.com.ph/deped-data/v1"
master <- read_parquet(file.path(base, "school_master_2023.parquet"))
panel <- read_parquet(file.path(base, "enrollment_panel_long.parquet"))

master |>
  filter(is_public, offers_es) |>
  group_by(region_const) |>
  summarise(median_ptr = median(ptr, na.rm = TRUE))

Command line. Run curl -O https://data.nash.com.ph/deped-data/v1/school_master_2023.csv.gz, then run sha256sum -c CHECKSUMS.sha256 in the same folder to verify what you downloaded.

What the data covers

The three assessment files are a selected set. They cover 44 of the 63 divisions in five regions (CAR, I, III, VIII, and IX) and about half of the public schools in those divisions. The schools were chosen by a rule DepEd has not published. Nothing in them describes learning outcomes elsewhere. The input modules are blank for private, SUC/LUC (state or local university or college), and overseas schools, because DepEd collects them for public schools only. These modules are personnel, facilities, water, sanitation, and hygiene (WASH), and information and communication technology (ICT). The enrollment files cover every school in the Learner Information System (LIS) each year, in all sectors.

How the data was cleaned

Every rule below is applied in code, never by hand, and is counted in the data quality report. The raw files were never edited.

1. Encoding and headers. The SY 2025-26 enrollment file mixes UTF-8 and Windows-1252 text and uses a different header scheme. It is decoded, and every year is read into one set of column names.
2. Labels harmonized. Whitespace is stripped, and SY 2025-26’s sector, management, and annex labels and its Yes/No flags are mapped to the coding the other eight years use. Division renames are mapped to the current name, and the CRLA file’s language spellings to the ELLNA technical-notes spelling. The full list is in harmonisation_rules.csv and in the table below.
3. A constant-boundary region. region_const places every school-year row in its SY 2025-26 region. The ARMM to BARMM change, the Cotabato City and Sulu transfers, and the new Negros Island Region therefore do not break a time series. The published label is kept beside it.
4. Blanks and zeros. A blank grade cell is 0 only when the school does not offer that level; a blank for a level it does offer is missing. SY 2025-26 writes blanks where earlier years wrote zeros. The 128 offered-but-blank levels in reporting schools are therefore read as 0, and the 46 schools that reported no enrollment at all are flagged as enrol_missing with blank totals. Nothing is imputed.
5. Public-only modules. Personnel, facilities, WASH, and ICT rows for non-public schools are blank and stay blank. The *_has_data columns say whether a school has any data in a module.
6. Implausible values flagged, not altered. The files have no negative counts, no scores outside 0 to 100, and no CRLA totals that fail to add up. Schools reporting more teachers or computers than learners are flagged. The only capped value is the derived comp_acad_w, which winsorizes academic computer counts at the 99th percentile because one school reports 15,699 tablets. The raw count is kept in comp_acad_total.
7. School IDs never re-linked. The 146 schools that appear to change ID are listed in likely_id_changes.csv. The panel counts each as an exit plus an entry, because DepEd publishes no mapping.
8. Types made consistent. Counts are integers (blank when missing), flags are true/false, shares are proportions from 0 to 1, and the same column has the same type in CSV and Parquet.

Harmonization rules

File-by-file guide to the columns

Columns are grouped as in codebook.csv. “Non-blank rows” is the number of rows with a value, which shows where a column applies; for example, personnel columns are filled for public schools only.

School master file, SY 2023-24

This file has one row per school in DepEd’s SY 2023-24 files: 60,167 schools in all sectors. Each row holds geography, enrollment by grade and sex, learner types, teachers, classrooms and buildings, toilets, water, electricity and internet, and computers and ICT equipment. It also holds the NAT, ELLNA, and CRLA results where they exist, and the derived ratios the report uses. Input modules are blank for non-public schools. Join the other SY 2023-24 files on school_id; the learners, personnel, facilities, ICT, and WASH files all contain the same 60,167 IDs.

Identification

Geography

Classification

Enrollment

Learner types

Personnel

Facilities

WASH

ICT

Assessment: NAT

Assessment: ELLNA

Assessment: CRLA

Assessment coverage

Derived indicators

Enrollment panel, SY 2017-18 to SY 2025-26

This file has one row per school for each school year in which it appears: 547,425 rows. Each row holds enrollment by grade, level, sex, and senior-high strand, with harmonized labels, constant-boundary regions, and entry, exit, and year-on-year columns. The CSV is split by school year; the Parquet file holds all nine years. A school-year is one row, so a school present in every year has nine rows.

Identification

Geography

Classification

Time

Enrollment

Data quality

Panel

School spine

This file has one row per school ID ever observed in the nine enrollment files: 66,967 IDs. Each row holds the first and last year, gaps, entry and exit flags, and the sector and location in the last year present. Use it to pick a balanced panel (present_all_years) or to see when a school entered or left the files.

Identification

Panel

Classification

Geography

CRLA reading levels, SY 2025-26

This file has one row per assessed school: 6,121 schools. Each row holds the learners at each CRLA reading level by grade and language, the school-level shares as proportions, and the SY 2025-26 Grade 1 to 3 enrollment. The source README calls the share columns percentages; they are proportions from 0 to 1 and are released as such.

Identification

Assessment: CRLA

Geography

Enrollment

Classification

Shrunk NAT and ELLNA school means

This file has one row per school with a NAT or ELLNA result: 6,840 schools. Each row holds the published school mean, the number of test-takers, and the empirical-Bayes estimate shrunk toward the division mean, for any school-level use. A NAT school mean rests on a median of six sampled learners, so its 95 percent margin is about six points. The shrunk estimate pulls small-sample schools toward their division mean and is the one to use for any school-level comparison. The guide How to Read DepEd Data explains the method.

Identification

Geography

Assessment: NAT

Derived indicators

Assessment: ELLNA

Helper tables

likely_id_changes.csv. This table lists 146 cases where a school that leaves the files reappears the next year under a new ID with the same name, municipality, and barangay. DepEd publishes no mapping, so the panel counts each as an exit plus an entry; this table lets you link them.

assessment_coverage_by_division.csv. For each of the 63 divisions in the five assessed regions, this table gives how many public schools with Grade 3 or Grade 6 learners have NAT, ELLNA, and CRLA results, and the share covered. Nineteen of the 63 divisions have no assessed schools.

Before you use the numbers

• Test results describe 44 divisions, not the country. State that in anything you publish. Do not compare NAT or ELLNA means across years, because the technical notes say difficulty is not equated.
• Do not rank schools on a six-learner mean. Use school_assessment_shrunk and report the number of test-takers beside any school figure.
• Ratios are undefined where the denominator is zero. 139 public schools report learners but no teachers, and 267 report no instructional classroom. Their ptr and pcr are blank, not infinite.
• Annex schools are separate rows. The technical notes compute pupil-teacher ratios with annexes merged into their mother schools; no mapping is published, so they are not merged here.
• Division names repeat. San Carlos City and San Fernando City each exist in two regions; group by region and division together.
• A school’s region can change. Use region_const for time series and region to match a year’s published tables.

Education Data, Transparency and Research covers DepEd’s own rules for using its assessment data and what EDCOM II found about DepEd’s data systems.

Terms of use and citation

Two sets of terms apply to this release.

The data files. The five cleaned datasets and the two helper tables are built from DepEd’s public datasets, which DepEd publishes under its Terms of Use and Fair Use Policy. Those terms require anything built on DepEd’s data to be distributed under the same terms, and they do not allow added restrictions, so the data files are distributed under DepEd’s terms, not ours. If you use them:

• credit the Department of Education as the source of the data;
• do not suggest that DepEd is involved in, associated with, or endorses your work;
• follow DepEd’s guidelines on assessment data: do not compare test scores across school years, and do not rank schools;
• read DepEd’s full terms before you publish anything built on the files.

DepEd’s terms also say its materials may not be modified beyond editorial changes without its approval. The original DepEd files were never edited, and every change in this release is listed in the harmonization rules and the codebook. If DepEd asks us to change or withdraw the release, we will.

Our own work. The cleaning pipeline, the codebook, the harmonization rules, the README and this documentation are © 2026 Mark Anthony Llego, NASH Reviewer. All rights reserved. To reuse, republish or adapt any of them, email support@nash.com.ph for written permission first, saying who you are, what you want to use and how.

NASH Reviewer is not affiliated with DepEd.

Suggested citation:

Mark Anthony Llego (2026). Cleaned DepEd school-level datasets, SY 2017-18 to SY 2025-26 (version 1.0) [data set]. NASH Reviewer. https://nash.com.ph/blog/download-cleaned-deped-school-data. Source data: Department of Education (Philippines) public datasets.

If you find an error, send us a note through the contact page with the file, column, and school_id. Corrections go into the next version with a note here.

Versions

• 1.0 (2026-10-25). First release: five datasets, two helper tables, codebook, harmonization rules, checksums, and README. Built from the enrollment files for SY 2017-18 to SY 2025-26, the SY 2023-24 learner, personnel, facilities, ICT, and WASH files, the SY 2023-24 NAT and ELLNA files, and the SY 2025-26 CRLA file.

Frequently asked questions

Is this official DepEd data?

The figures are DepEd’s, from its publicly released datasets, but the cleaned files are not an official DepEd release. The cleaning, the harmonized labels, and the derived columns are ours and are documented column by column, so you can trace any value to the source file. For official figures, cite DepEd’s own releases.

Can I use these DepEd data files in a thesis, a report, or an app?

Yes, under DepEd’s terms of use: credit DepEd as the source, do not suggest DepEd’s involvement, and follow its rules on assessment data (no comparison of test scores across school years, no ranking of schools). Please also cite this release. To reuse our cleaning code, codebook or documentation, email support@nash.com.ph for permission first. The terms of use have the details.

Why are so many cells blank in the DepEd school data?

There are three reasons, and the codebook’s non-blank counts show which one applies. The variable may not apply to the school, such as a level it does not offer or a public-only module for a private school. The variable may not have existed that year, such as the shs_unique column (unique or other senior-high programs) before SY 2024-25 or districts in SY 2017-18. Or the school did not report, as 46 schools in SY 2025-26 did not. Blanks are never zeros.

Why are the files in both CSV and Parquet?

CSV opens anywhere but loses types and compresses less well. Parquet keeps integers, booleans, and text as they are, loads in seconds in pandas, R, or DuckDB, and is a third smaller. The values are identical.

How do I follow one DepEd school across school years?

Filter the enrollment panel on school_id, or start from the school spine, which has one row per ID with its first and last year. Check likely_id_changes.csv for the 146 schools that seem to have changed ID.

Can I rank schools by NAT score with these files?

No, not on the published means. Each rests on a median of six sampled learners, and a six-point difference is within noise. Use the shrunk means and show the number of test-takers. The guide How to Read DepEd Data explains why.

Will the cleaned DepEd data be updated?

Yes, when DepEd releases new files. We will rebuild the datasets as a new version folder and list the changes here. The pipeline that produced this release is deterministic, so the checksums above will reproduce.

More in this series

• DepEd Enrollment Trends SY 2017-2018 to SY 2025-2026
• Enrollment by school year: SY 2017-18, SY 2018-19, SY 2019-20, SY 2020-21, SY 2021-22, SY 2022-23, SY 2023-24, SY 2024-25, SY 2025-26
• DepEd Learner Data SY 2023-2024: IP, ALIVE and School Size
• DepEd Teachers and School Personnel Data, SY 2023-2024
• DepEd Classroom and School Building Data, SY 2023-2024
• DepEd WASH Data SY 2023-2024: Toilets, Water, Electricity
• DepEd ICT Data SY 2023-2024: Computers and Internet
• NAT Grade 6 Results SY 2023-2024: DepEd Division Scores
• ELLNA Grade 3 Results SY 2023-2024: DepEd Division Scores
• CRLA Results SY 2025-2026: Reading Levels in Grades 1 to 3
• DepEd School Inputs vs Test Results: Multigrade and Internet
• DepEd Open Data Quality: Errors, Blanks and Coverage Limits
• The full report: Philippine basic education in numbers

Guides for aspiring school heads

• DepEd Data for Aspiring School Heads: Where to Start
• NASH Practice Questions With Answers: 50 Items on the PPSSH
• How to Read DepEd Data: A Guide for Aspiring School Heads
• EDCOM II Final Report: Guide for Aspiring School Heads
• National Education Plan 2026-2035: Guide for School Heads
• NASH Study Guide: DepEd Memorandum 069, s. 2025 and PPSSH
• PPSSH Domain 1: Leading Strategically With DepEd Data
• PPSSH Domain 2: Managing School Operations With DepEd Data
• PPSSH Domain 3: Teaching and Learning With DepEd Data
• PPSSH Domain 4: Developing Self and Others With DepEd Data
• PPSSH Domain 5: Building Connections With DepEd Data

Data preparation and documentation by Mark Anthony Llego for NASH Reviewer. Charts in the series carry the same credit.