Add model card, model specs, indicator config and the synthetic live nowcast
Browse files- LICENSE +21 -0
- README.md +142 -0
- config/indicators.yaml +319 -0
- live_nowcast.json +490 -0
- model_specs.json +387 -0
LICENSE
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
MIT License
|
| 2 |
+
|
| 3 |
+
Copyright (c) 2026 Thanatchon Chongmankhong
|
| 4 |
+
|
| 5 |
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
| 6 |
+
of this software and associated documentation files (the "Software"), to deal
|
| 7 |
+
in the Software without restriction, including without limitation the rights
|
| 8 |
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
| 9 |
+
copies of the Software, and to permit persons to whom the Software is
|
| 10 |
+
furnished to do so, subject to the following conditions:
|
| 11 |
+
|
| 12 |
+
The above copyright notice and this permission notice shall be included in all
|
| 13 |
+
copies or substantial portions of the Software.
|
| 14 |
+
|
| 15 |
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
| 16 |
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
| 17 |
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
| 18 |
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
| 19 |
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
| 20 |
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
| 21 |
+
SOFTWARE.
|
README.md
ADDED
|
@@ -0,0 +1,142 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
- th
|
| 5 |
+
license: mit
|
| 6 |
+
library_name: scikit-learn
|
| 7 |
+
pipeline_tag: time-series-forecasting
|
| 8 |
+
datasets:
|
| 9 |
+
- thanatchon36/thainowcast
|
| 10 |
+
tags:
|
| 11 |
+
- nowcasting
|
| 12 |
+
- thailand
|
| 13 |
+
- macroeconomics
|
| 14 |
+
- gdp
|
| 15 |
+
- inflation
|
| 16 |
+
- dynamic-factor-model
|
| 17 |
+
- midas
|
| 18 |
+
- elastic-net
|
| 19 |
+
- lightgbm
|
| 20 |
+
- synthetic
|
| 21 |
+
---
|
| 22 |
+
|
| 23 |
+
# thainowcast: nowcasting models for Thai GDP and inflation
|
| 24 |
+
|
| 25 |
+
> **No weights here, and no results about Thailand.** These models keep no fitted state. At every forecast
|
| 26 |
+
> origin each one is re-estimated from scratch on the data released by that date, so a frozen set of weights
|
| 27 |
+
> would be stale the day after it was saved. This repository records the models instead: the classes,
|
| 28 |
+
> hyperparameters, feature sets and the model × feature-set grid. Every number quoted below comes from
|
| 29 |
+
> **synthetic test data**. It shows that the pipeline works, not how well these models nowcast the Thai economy.
|
| 30 |
+
|
| 31 |
+
These are the models of the working paper *ThaiNowcast: Pseudo-Real-Time Nowcasting of Thai GDP and Inflation
|
| 32 |
+
with Official and Alternative Data* by Thanatchon Chongmankhong
|
| 33 |
+
([read online](https://thanatchon36.github.io/thainowcast/paper/) ·
|
| 34 |
+
[PDF](https://github.com/thanatchon36/thainowcast/raw/main/paper/paper.pdf) ·
|
| 35 |
+
[DOI 10.5281/zenodo.23995547](https://doi.org/10.5281/zenodo.23995547)).
|
| 36 |
+
Code: <https://github.com/thanatchon36/thainowcast>. Data and full results:
|
| 37 |
+
[`thanatchon36/thainowcast`](https://huggingface.co/datasets/thanatchon36/thainowcast) (dataset).
|
| 38 |
+
|
| 39 |
+
*Independent personal research. Does not represent the views of any employer or institution.*
|
| 40 |
+
|
| 41 |
+
## Targets
|
| 42 |
+
|
| 43 |
+
| target | frequency | forecast origins |
|
| 44 |
+
|---|---|---|
|
| 45 |
+
| `gdp_yoy`: real GDP growth, y-o-y (%), NESDC | quarterly | M1, M2, M3 (end of each month of the quarter), PR (the day before the official release) |
|
| 46 |
+
| `cpi_yoy`: headline CPI inflation, y-o-y (%), TPSO | monthly | M0 (mid-month), PR |
|
| 47 |
+
|
| 48 |
+
## Models
|
| 49 |
+
|
| 50 |
+
| code | class | hyperparameters |
|
| 51 |
+
|---|---|---|
|
| 52 |
+
| `rw` | `RandomWalk` | last released value |
|
| 53 |
+
| `ar` | `AR` | AR(p), p ≤ 4 chosen by AIC |
|
| 54 |
+
| `bridge` | `Bridge` | indicators completed to the end of the target period with AR(1) forecasts, averaged to the target frequency, 2 principal components, OLS |
|
| 55 |
+
| `midas` | `MIDAS` | one regressor (mean of standardized indicators), 12 monthly lags, exponential Almon weights re-estimated for each number of missing months. GDP only |
|
| 56 |
+
| `dfm` | `DFM` | mixed-frequency dynamic factor model (statsmodels `DynamicFactorMQ`), 1 factor, AR(1) factor, EM up to 50 iterations, re-estimated once a year; also gives the news decomposition |
|
| 57 |
+
| `enet` | `ElasticNetNowcast` | `ElasticNetCV`, l1 ratio in {0.1, 0.5, 0.9}, 40 alphas, time-series CV with up to 4 folds |
|
| 58 |
+
| `lgbm` | `LightGBMNowcast` | 300 trees, learning rate 0.03, 4 leaves, min 5 samples per leaf, row and column subsampling 0.8, L2 1.0, deterministic |
|
| 59 |
+
|
| 60 |
+
Feature sets: `hard_only` (14 official monthly statistics), `hard_trends`, `hard_lights`, `hard_no2` and
|
| 61 |
+
`hard_prices` (hard data plus one family of alternative data: Google Trends, nighttime lights, Sentinel-5P NO2,
|
| 62 |
+
online prices), `all` (hard data plus all five families, mobility included) and `none` for the univariate
|
| 63 |
+
benchmarks. GDP runs 20 model × feature-set pairs and CPI 22. Each model's 80% interval is the conformal quantile
|
| 64 |
+
of its absolute backtest errors at the same origin.
|
| 65 |
+
|
| 66 |
+
## Files
|
| 67 |
+
|
| 68 |
+
| path | content |
|
| 69 |
+
|---|---|
|
| 70 |
+
| `model_specs.json` | every model's class and hyperparameters, the indicators in each feature set, the backtest settings and the full grid per target, pinned to code commit `ce9dd21` |
|
| 71 |
+
| `config/indicators.yaml` | the indicator registry: agency, frequency, start, publication lag, transform and the source of each series |
|
| 72 |
+
| `live_nowcast.json` | **synthetic.** The live nowcast as of 2026-11-05: every model's value and interval, the headline model and the GDP news decomposition. Rerunning the snippet under Use reproduces the GDP headline, 2.622 |
|
| 73 |
+
| `LICENSE` | MIT, as for the code |
|
| 74 |
+
|
| 75 |
+
## Accuracy on synthetic data
|
| 76 |
+
|
| 77 |
+
RMSFE at the last origin (PR) over the full backtest, relative to the AR benchmark (below 1 beats AR).
|
| 78 |
+
GDP covers 2015Q1 to 2026Q2 (46 quarters), CPI 2015-01 to 2026-09 (141 months).
|
| 79 |
+
|
| 80 |
+
| model | GDP `hard_only` | GDP `all` | CPI `hard_only` | CPI `all` |
|
| 81 |
+
|---|---|---|---|---|
|
| 82 |
+
| `enet` | **0.46** | 0.78 | 0.68 | **0.54** |
|
| 83 |
+
| `bridge` | 0.55 | 0.66 | 1.45 | 1.29 |
|
| 84 |
+
| `dfm` | 0.64 | 0.89 | 0.92 | 0.92 |
|
| 85 |
+
| `midas` | 0.64 | 0.66 | – | – |
|
| 86 |
+
| `lgbm` | 0.73 | 0.72 | 0.72 | 0.60 |
|
| 87 |
+
| `rw` | 1.04 | | 1.02 | |
|
| 88 |
+
|
| 89 |
+
The synthetic generator sets how much each data family helps, so this ranking tells you nothing about Thailand.
|
| 90 |
+
The full tables, Diebold-Mariano tests, regime splits and ablations are in the
|
| 91 |
+
[dataset](https://huggingface.co/datasets/thanatchon36/thainowcast/tree/main/results).
|
| 92 |
+
|
| 93 |
+
## Use
|
| 94 |
+
|
| 95 |
+
The models run inside the pipeline, so you use the code rather than a download from this repository:
|
| 96 |
+
|
| 97 |
+
```bash
|
| 98 |
+
git clone https://github.com/thanatchon36/thainowcast && cd thainowcast
|
| 99 |
+
make setup # uv venv + pip install -e ".[dev,app]"
|
| 100 |
+
make all # data -> eval (backtest) -> live -> dashboard -> paper
|
| 101 |
+
```
|
| 102 |
+
|
| 103 |
+
To nowcast for real, put the official CSVs named in
|
| 104 |
+
[`docs/data_sources.md`](https://github.com/thanatchon36/thainowcast/blob/main/docs/data_sources.md) into
|
| 105 |
+
`data/raw/` (the `synthetic_*` files are used only when a real one is missing, with a warning), then run
|
| 106 |
+
`make all ASOF=YYYY-MM-DD`. One model on its own:
|
| 107 |
+
|
| 108 |
+
```python
|
| 109 |
+
from thainowcast.config import load_config
|
| 110 |
+
from thainowcast.data.loaders import load_panel
|
| 111 |
+
from thainowcast.features import transform_wide
|
| 112 |
+
from thainowcast.models import MODELS
|
| 113 |
+
from thainowcast.config import RAW_DIR
|
| 114 |
+
import pandas as pd
|
| 115 |
+
|
| 116 |
+
cfg, asof = load_config(), pd.Timestamp("2026-11-05")
|
| 117 |
+
panel = load_panel(cfg, RAW_DIR)
|
| 118 |
+
y = panel.series_available(asof, "gdp_yoy") # GDP released by asof
|
| 119 |
+
X = transform_wide(panel.wide_available(asof, list(cfg.indicators), "M"), cfg.indicators)
|
| 120 |
+
X = X[cfg.set_columns("hard_only")]
|
| 121 |
+
print(MODELS["enet"]().nowcast(y, X, pd.Period("2026Q3", "Q"), asof))
|
| 122 |
+
```
|
| 123 |
+
|
| 124 |
+
## Limitations
|
| 125 |
+
|
| 126 |
+
- Trained and evaluated on synthetic data only. Real Thai series have revisions, breaks and gaps the generator
|
| 127 |
+
does not reproduce, and the backtest uses first releases, not real-time vintages.
|
| 128 |
+
- Publication lags in `config/indicators.yaml` are approximations of each agency's usual timing.
|
| 129 |
+
- The panel is short: 46 GDP quarters in the backtest, so differences between models are noisy.
|
| 130 |
+
|
| 131 |
+
## Citation
|
| 132 |
+
|
| 133 |
+
```bibtex
|
| 134 |
+
@misc{chongmankhong2026thainowcast,
|
| 135 |
+
author = {Chongmankhong, Thanatchon},
|
| 136 |
+
title = {ThaiNowcast: Pseudo-Real-Time Nowcasting of Thai GDP and Inflation with Official and Alternative Data},
|
| 137 |
+
year = {2026},
|
| 138 |
+
doi = {10.5281/zenodo.23995547},
|
| 139 |
+
url = {https://thanatchon36.github.io/thainowcast/paper/},
|
| 140 |
+
note = {Working paper; results are from synthetic test data}
|
| 141 |
+
}
|
| 142 |
+
```
|
config/indicators.yaml
ADDED
|
@@ -0,0 +1,319 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# ThaiNowcast — indicator registry and release calendar
|
| 2 |
+
# ---------------------------------------------------------------------------
|
| 3 |
+
# start : first period of the LEVEL series (a YoY feature needs 12 more months)
|
| 4 |
+
# end : last period, only for discontinued series
|
| 5 |
+
# release_lag_days : typical calendar days between the end of the reference period and
|
| 6 |
+
# publication. Approximations taken from each agency's usual release
|
| 7 |
+
# timing — verify against the official calendars before a real run [ต้องตรวจ].
|
| 8 |
+
# transform : yoy | log_yoy | yoy_diff | level
|
| 9 |
+
# file : CSV expected in data/raw/ with columns `period,value`.
|
| 10 |
+
# `synthetic_<file>` is used ONLY when the real file is missing, with a warning.
|
| 11 |
+
# fetch : where the real file comes from (see docs/data_sources.md)
|
| 12 |
+
|
| 13 |
+
targets:
|
| 14 |
+
gdp_yoy:
|
| 15 |
+
name_th: "GDP จริง YoY (%)"
|
| 16 |
+
name_en: "Real GDP growth, y-o-y (%)"
|
| 17 |
+
agency: "NESDC (สภาพัฒน์)"
|
| 18 |
+
frequency: Q
|
| 19 |
+
start: "2005Q1"
|
| 20 |
+
release_lag_days: 48
|
| 21 |
+
transform: level
|
| 22 |
+
fetch: "manual: nesdc.go.th > QGDP press release > Excel table, YoY % of GDP at chained volume"
|
| 23 |
+
file: gdp_yoy.csv
|
| 24 |
+
gdp_qoq_sa:
|
| 25 |
+
name_th: "GDP จริง QoQ SA (%)"
|
| 26 |
+
name_en: "Real GDP growth, q-o-q seasonally adjusted (%)"
|
| 27 |
+
agency: "NESDC (สภาพัฒน์)"
|
| 28 |
+
frequency: Q
|
| 29 |
+
start: "2005Q1"
|
| 30 |
+
release_lag_days: 48
|
| 31 |
+
transform: level
|
| 32 |
+
fetch: "manual: same NESDC release as gdp_yoy, seasonally adjusted table"
|
| 33 |
+
file: gdp_qoq_sa.csv
|
| 34 |
+
cpi_yoy:
|
| 35 |
+
name_th: "เงินเฟ้อทั่วไป YoY (%)"
|
| 36 |
+
name_en: "Headline CPI inflation, y-o-y (%)"
|
| 37 |
+
agency: "TPSO กระทรวงพาณิชย์"
|
| 38 |
+
frequency: M
|
| 39 |
+
start: "2005-01"
|
| 40 |
+
release_lag_days: 6
|
| 41 |
+
transform: level
|
| 42 |
+
fetch: "manual: tpso.go.th / price.moc.go.th > CPI monthly report > headline YoY"
|
| 43 |
+
file: cpi_yoy.csv
|
| 44 |
+
|
| 45 |
+
indicators:
|
| 46 |
+
# ---------------- hard data: official monthly statistics ----------------
|
| 47 |
+
mpi:
|
| 48 |
+
name_th: "ดัชนีผลผลิตอุตสาหกรรม"
|
| 49 |
+
name_en: "Manufacturing Production Index"
|
| 50 |
+
agency: "OIE (สศอ.)"
|
| 51 |
+
group: hard
|
| 52 |
+
frequency: M
|
| 53 |
+
start: "2004-01"
|
| 54 |
+
release_lag_days: 30
|
| 55 |
+
transform: yoy
|
| 56 |
+
fetch: "manual: oie.go.th > สถิติอุตสาหกรรม > MPI xlsx (or BOT API EC_EI_027) [ต้องตรวจ]"
|
| 57 |
+
file: mpi.csv
|
| 58 |
+
pci:
|
| 59 |
+
name_th: "ดัชนีการบริโภคภาคเอกชน (PCI)"
|
| 60 |
+
name_en: "Private Consumption Index"
|
| 61 |
+
agency: "BOT (ธปท.)"
|
| 62 |
+
group: hard
|
| 63 |
+
frequency: M
|
| 64 |
+
start: "2004-01"
|
| 65 |
+
release_lag_days: 30
|
| 66 |
+
transform: yoy
|
| 67 |
+
fetch: "api: BOT API portal (apiportal.bot.or.th), Economic & Financial statistics, PCI [ต้องตรวจ series code]"
|
| 68 |
+
file: pci.csv
|
| 69 |
+
pii:
|
| 70 |
+
name_th: "ดัชนีการลงทุนภาคเอกชน (PII)"
|
| 71 |
+
name_en: "Private Investment Index"
|
| 72 |
+
agency: "BOT (ธปท.)"
|
| 73 |
+
group: hard
|
| 74 |
+
frequency: M
|
| 75 |
+
start: "2004-01"
|
| 76 |
+
release_lag_days: 30
|
| 77 |
+
transform: yoy
|
| 78 |
+
fetch: "api: BOT API portal, PII [ต้องตรวจ series code]"
|
| 79 |
+
file: pii.csv
|
| 80 |
+
credit:
|
| 81 |
+
name_th: "สินเชื่อระบบธนาคารพาณิชย์"
|
| 82 |
+
name_en: "Commercial bank loans outstanding"
|
| 83 |
+
agency: "BOT (ธปท.)"
|
| 84 |
+
group: hard
|
| 85 |
+
frequency: M
|
| 86 |
+
start: "2004-01"
|
| 87 |
+
release_lag_days: 30
|
| 88 |
+
transform: yoy
|
| 89 |
+
fetch: "api: BOT API portal, loans outstanding by sector"
|
| 90 |
+
file: credit.csv
|
| 91 |
+
promptpay:
|
| 92 |
+
name_th: "มูลค่าธุรกรรมพร้อมเพย์"
|
| 93 |
+
name_en: "PromptPay transaction value"
|
| 94 |
+
agency: "BOT (ธปท.)"
|
| 95 |
+
group: hard
|
| 96 |
+
frequency: M
|
| 97 |
+
start: "2016-01"
|
| 98 |
+
release_lag_days: 30
|
| 99 |
+
transform: yoy
|
| 100 |
+
fetch: "manual: BOT payment systems statistics (bot.or.th > Payment Systems > Statistics)"
|
| 101 |
+
file: promptpay.csv
|
| 102 |
+
exports:
|
| 103 |
+
name_th: "มูลค่าส่งออก (USD)"
|
| 104 |
+
name_en: "Merchandise exports (USD)"
|
| 105 |
+
agency: "Customs Department (กรมศุลกากร)"
|
| 106 |
+
group: hard
|
| 107 |
+
frequency: M
|
| 108 |
+
start: "2004-01"
|
| 109 |
+
release_lag_days: 25
|
| 110 |
+
transform: yoy
|
| 111 |
+
fetch: "manual: customs.go.th > สถิติการค้า; also MOC press release"
|
| 112 |
+
file: exports.csv
|
| 113 |
+
imports:
|
| 114 |
+
name_th: "มูลค่านำเข้า (USD)"
|
| 115 |
+
name_en: "Merchandise imports (USD)"
|
| 116 |
+
agency: "Customs Department (กรมศุลกากร)"
|
| 117 |
+
group: hard
|
| 118 |
+
frequency: M
|
| 119 |
+
start: "2004-01"
|
| 120 |
+
release_lag_days: 25
|
| 121 |
+
transform: yoy
|
| 122 |
+
fetch: "manual: customs.go.th > สถิติการค้า"
|
| 123 |
+
file: imports.csv
|
| 124 |
+
tourists:
|
| 125 |
+
name_th: "นักท่องเที่ยวต่างชาติ (คน)"
|
| 126 |
+
name_en: "International tourist arrivals"
|
| 127 |
+
agency: "MOTS (กระทรวงการท่องเที่ยวฯ)"
|
| 128 |
+
group: hard
|
| 129 |
+
frequency: M
|
| 130 |
+
start: "2004-01"
|
| 131 |
+
release_lag_days: 20
|
| 132 |
+
transform: log_yoy
|
| 133 |
+
fetch: "manual: mots.go.th > สถิตินักท่องเที่ยว > monthly arrivals"
|
| 134 |
+
file: tourists.csv
|
| 135 |
+
aot_pax:
|
| 136 |
+
name_th: "ผู้โดยสารสนามบิน AOT"
|
| 137 |
+
name_en: "AOT airport passengers"
|
| 138 |
+
agency: "AOT (ทอท.)"
|
| 139 |
+
group: hard
|
| 140 |
+
frequency: M
|
| 141 |
+
start: "2004-01"
|
| 142 |
+
release_lag_days: 20
|
| 143 |
+
transform: log_yoy
|
| 144 |
+
fetch: "manual: airportthai.co.th > Air Traffic Statistics"
|
| 145 |
+
file: aot_pax.csv
|
| 146 |
+
electricity:
|
| 147 |
+
name_th: "ปริมาณการใช้ไฟฟ้า"
|
| 148 |
+
name_en: "Electricity consumption"
|
| 149 |
+
agency: "EPPO (สนพ.)"
|
| 150 |
+
group: hard
|
| 151 |
+
frequency: M
|
| 152 |
+
start: "2004-01"
|
| 153 |
+
release_lag_days: 45
|
| 154 |
+
transform: yoy
|
| 155 |
+
fetch: "manual: eppo.go.th > สถิติพลังงาน > electricity consumption by sector"
|
| 156 |
+
file: electricity.csv
|
| 157 |
+
vehicle_reg:
|
| 158 |
+
name_th: "ยอดจดทะเบียนรถใหม่"
|
| 159 |
+
name_en: "New vehicle registrations"
|
| 160 |
+
agency: "DLT (กรมการขนส่งทางบก)"
|
| 161 |
+
group: hard
|
| 162 |
+
frequency: M
|
| 163 |
+
start: "2004-01"
|
| 164 |
+
release_lag_days: 15
|
| 165 |
+
transform: yoy
|
| 166 |
+
fetch: "manual: web.dlt.go.th/statistics > จดทะเบียนใหม่"
|
| 167 |
+
file: vehicle_reg.csv
|
| 168 |
+
expressway:
|
| 169 |
+
name_th: "ปริมาณจราจรทางพิเศษ"
|
| 170 |
+
name_en: "Expressway traffic volume"
|
| 171 |
+
agency: "EXAT (กทพ.)"
|
| 172 |
+
group: hard
|
| 173 |
+
frequency: M
|
| 174 |
+
start: "2004-01"
|
| 175 |
+
release_lag_days: 30
|
| 176 |
+
transform: yoy
|
| 177 |
+
fetch: "manual: exat.co.th > สถิติปริมาณจราจร"
|
| 178 |
+
file: expressway.csv
|
| 179 |
+
oil_retail:
|
| 180 |
+
name_th: "ราคาขายปลีกดีเซล"
|
| 181 |
+
name_en: "Retail diesel price (monthly average)"
|
| 182 |
+
agency: "EPPO (สนพ.) / PTT"
|
| 183 |
+
group: hard
|
| 184 |
+
frequency: M
|
| 185 |
+
start: "2004-01"
|
| 186 |
+
release_lag_days: 1
|
| 187 |
+
transform: yoy
|
| 188 |
+
fetch: "scrape: eppo.go.th petroleum price statistics (daily, average to month)"
|
| 189 |
+
file: oil_retail.csv
|
| 190 |
+
agri_price:
|
| 191 |
+
name_th: "ดัชนีราคาสินค้าเกษตร"
|
| 192 |
+
name_en: "Farm price index"
|
| 193 |
+
agency: "OAE (สศก.)"
|
| 194 |
+
group: hard
|
| 195 |
+
frequency: M
|
| 196 |
+
start: "2004-01"
|
| 197 |
+
release_lag_days: 15
|
| 198 |
+
transform: yoy
|
| 199 |
+
fetch: "manual: oae.go.th > ดัชนีราคาสินค้าเกษตร"
|
| 200 |
+
file: agri_price.csv
|
| 201 |
+
|
| 202 |
+
# ---------------- alternative data ----------------
|
| 203 |
+
gt_unemploy:
|
| 204 |
+
name_th: "Google Trends: ว่างงาน / หางาน"
|
| 205 |
+
name_en: "Google Trends: unemployment / job search (TH)"
|
| 206 |
+
agency: "Google Trends"
|
| 207 |
+
group: alt_trends
|
| 208 |
+
frequency: M
|
| 209 |
+
start: "2008-01"
|
| 210 |
+
release_lag_days: 3
|
| 211 |
+
transform: yoy_diff
|
| 212 |
+
fetch: "api: pytrends, geo=TH, topic 'ว่างงาน','หางาน'; weekly -> monthly mean"
|
| 213 |
+
file: gt_unemploy.csv
|
| 214 |
+
gt_loan:
|
| 215 |
+
name_th: "Google Trends: ขอสินเชื่อ / กู้เงิน"
|
| 216 |
+
name_en: "Google Trends: loan / borrowing (TH)"
|
| 217 |
+
agency: "Google Trends"
|
| 218 |
+
group: alt_trends
|
| 219 |
+
frequency: M
|
| 220 |
+
start: "2008-01"
|
| 221 |
+
release_lag_days: 3
|
| 222 |
+
transform: yoy_diff
|
| 223 |
+
fetch: "api: pytrends, geo=TH, 'กู้เงิน','สินเชื่อ'"
|
| 224 |
+
file: gt_loan.csv
|
| 225 |
+
gt_car_installment:
|
| 226 |
+
name_th: "Google Trends: ผ่อนรถ"
|
| 227 |
+
name_en: "Google Trends: car installment (TH)"
|
| 228 |
+
agency: "Google Trends"
|
| 229 |
+
group: alt_trends
|
| 230 |
+
frequency: M
|
| 231 |
+
start: "2008-01"
|
| 232 |
+
release_lag_days: 3
|
| 233 |
+
transform: yoy_diff
|
| 234 |
+
fetch: "api: pytrends, geo=TH, 'ผ่อนรถ'"
|
| 235 |
+
file: gt_car_installment.csv
|
| 236 |
+
gt_travel:
|
| 237 |
+
name_th: "Google Trends: ท่องเที่ยว / จองโรงแรม"
|
| 238 |
+
name_en: "Google Trends: travel / hotel booking (TH)"
|
| 239 |
+
agency: "Google Trends"
|
| 240 |
+
group: alt_trends
|
| 241 |
+
frequency: M
|
| 242 |
+
start: "2008-01"
|
| 243 |
+
release_lag_days: 3
|
| 244 |
+
transform: yoy_diff
|
| 245 |
+
fetch: "api: pytrends, geo=TH, 'ที่พัก','จองโรงแรม'"
|
| 246 |
+
file: gt_travel.csv
|
| 247 |
+
gt_prices:
|
| 248 |
+
name_th: "Google Trends: ของแพง / ราคาน้ำมัน"
|
| 249 |
+
name_en: "Google Trends: prices / fuel price (TH)"
|
| 250 |
+
agency: "Google Trends"
|
| 251 |
+
group: alt_trends
|
| 252 |
+
frequency: M
|
| 253 |
+
start: "2008-01"
|
| 254 |
+
release_lag_days: 3
|
| 255 |
+
transform: yoy_diff
|
| 256 |
+
fetch: "api: pytrends, geo=TH, 'ของแพง','ราคาน้ำมันวันนี้'"
|
| 257 |
+
file: gt_prices.csv
|
| 258 |
+
nightlights:
|
| 259 |
+
name_th: "แสงไฟกลางคืน (VIIRS Black Marble)"
|
| 260 |
+
name_en: "Nighttime lights radiance, national sum (VNP46A3)"
|
| 261 |
+
agency: "NASA Black Marble"
|
| 262 |
+
group: alt_lights
|
| 263 |
+
frequency: M
|
| 264 |
+
start: "2012-04"
|
| 265 |
+
release_lag_days: 40
|
| 266 |
+
transform: log_yoy
|
| 267 |
+
fetch: "api: NASA Earthdata (ladsweb) VNP46A3 monthly tiles -> provincial zonal sum with GADM boundaries"
|
| 268 |
+
file: nightlights.csv
|
| 269 |
+
no2:
|
| 270 |
+
name_th: "NO2 โทรโพสเฟียร์ (Sentinel-5P)"
|
| 271 |
+
name_en: "Tropospheric NO2 column, industrial provinces (Sentinel-5P)"
|
| 272 |
+
agency: "ESA Copernicus"
|
| 273 |
+
group: alt_no2
|
| 274 |
+
frequency: M
|
| 275 |
+
start: "2018-07"
|
| 276 |
+
release_lag_days: 5
|
| 277 |
+
transform: log_yoy
|
| 278 |
+
fetch: "api: Google Earth Engine COPERNICUS/S5P/OFFL/L3_NO2, monthly mean over EEC + Samut Prakan"
|
| 279 |
+
file: no2.csv
|
| 280 |
+
online_prices:
|
| 281 |
+
name_th: "ดัชนีราคาสินค้าออนไลน์"
|
| 282 |
+
name_en: "Online price index (scraped daily, Billion-Prices style)"
|
| 283 |
+
agency: "own scraper"
|
| 284 |
+
group: alt_prices
|
| 285 |
+
frequency: M
|
| 286 |
+
start: "2015-01"
|
| 287 |
+
release_lag_days: 1
|
| 288 |
+
transform: yoy
|
| 289 |
+
fetch: "scrape: own daily scraper of major online grocers (check ToS) [ต้องตรวจ]"
|
| 290 |
+
file: online_prices.csv
|
| 291 |
+
mobility:
|
| 292 |
+
name_th: "Google Mobility: retail & recreation"
|
| 293 |
+
name_en: "Google Community Mobility: retail & recreation (% vs baseline)"
|
| 294 |
+
agency: "Google (discontinued 2022-10)"
|
| 295 |
+
group: alt_mobility
|
| 296 |
+
frequency: M
|
| 297 |
+
start: "2020-02"
|
| 298 |
+
end: "2022-10"
|
| 299 |
+
release_lag_days: 2
|
| 300 |
+
transform: level
|
| 301 |
+
fetch: "manual: google.com/covid19/mobility archived CSV, TH national"
|
| 302 |
+
file: mobility.csv
|
| 303 |
+
|
| 304 |
+
feature_sets:
|
| 305 |
+
hard_only: [hard]
|
| 306 |
+
hard_trends: [hard, alt_trends]
|
| 307 |
+
hard_lights: [hard, alt_lights]
|
| 308 |
+
hard_no2: [hard, alt_no2]
|
| 309 |
+
hard_prices: [hard, alt_prices]
|
| 310 |
+
all: [hard, alt_trends, alt_lights, alt_no2, alt_prices, alt_mobility]
|
| 311 |
+
|
| 312 |
+
backtest:
|
| 313 |
+
start: "2015-01"
|
| 314 |
+
benchmark: ar
|
| 315 |
+
conformal_alpha: 0.2
|
| 316 |
+
regimes:
|
| 317 |
+
pre_covid: ["2015-01", "2019-12"]
|
| 318 |
+
covid: ["2020-01", "2021-12"]
|
| 319 |
+
post_covid: ["2022-01", "2035-12"]
|
live_nowcast.json
ADDED
|
@@ -0,0 +1,490 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"asof": "2026-11-05",
|
| 3 |
+
"synthetic": true,
|
| 4 |
+
"synthetic_ids": [
|
| 5 |
+
"agri_price",
|
| 6 |
+
"aot_pax",
|
| 7 |
+
"cpi_yoy",
|
| 8 |
+
"credit",
|
| 9 |
+
"electricity",
|
| 10 |
+
"exports",
|
| 11 |
+
"expressway",
|
| 12 |
+
"gdp_qoq_sa",
|
| 13 |
+
"gdp_yoy",
|
| 14 |
+
"gt_car_installment",
|
| 15 |
+
"gt_loan",
|
| 16 |
+
"gt_prices",
|
| 17 |
+
"gt_travel",
|
| 18 |
+
"gt_unemploy",
|
| 19 |
+
"imports",
|
| 20 |
+
"mobility",
|
| 21 |
+
"mpi",
|
| 22 |
+
"nightlights",
|
| 23 |
+
"no2",
|
| 24 |
+
"oil_retail",
|
| 25 |
+
"online_prices",
|
| 26 |
+
"pci",
|
| 27 |
+
"pii",
|
| 28 |
+
"promptpay",
|
| 29 |
+
"tourists",
|
| 30 |
+
"vehicle_reg"
|
| 31 |
+
],
|
| 32 |
+
"targets": {
|
| 33 |
+
"gdp_yoy": {
|
| 34 |
+
"label": "Real GDP growth (y-o-y)",
|
| 35 |
+
"period": "2026Q3",
|
| 36 |
+
"release_expected": "2026-11-17",
|
| 37 |
+
"headline": {
|
| 38 |
+
"model": "enet",
|
| 39 |
+
"feature_set": "hard_only",
|
| 40 |
+
"value": 2.622081984570472,
|
| 41 |
+
"lo": 1.2321626043496732,
|
| 42 |
+
"hi": 4.012001364791271,
|
| 43 |
+
"rmsfe_full": 1.1729679933805253
|
| 44 |
+
},
|
| 45 |
+
"models": [
|
| 46 |
+
{
|
| 47 |
+
"model": "rw",
|
| 48 |
+
"feature_set": "none",
|
| 49 |
+
"value": 4.035735,
|
| 50 |
+
"lo": 2.1696509999999996,
|
| 51 |
+
"hi": 5.901819
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"model": "ar",
|
| 55 |
+
"feature_set": "none",
|
| 56 |
+
"value": 3.7247021640278297,
|
| 57 |
+
"lo": 1.6627235019632787,
|
| 58 |
+
"hi": 5.786680826092381
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"model": "bridge",
|
| 62 |
+
"feature_set": "hard_only",
|
| 63 |
+
"value": 2.744188700020815,
|
| 64 |
+
"lo": 1.1611176888087378,
|
| 65 |
+
"hi": 4.327259711232892
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"model": "bridge",
|
| 69 |
+
"feature_set": "hard_trends",
|
| 70 |
+
"value": 2.674056418037423,
|
| 71 |
+
"lo": 1.2520382531462815,
|
| 72 |
+
"hi": 4.096074582928565
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"model": "bridge",
|
| 76 |
+
"feature_set": "hard_lights",
|
| 77 |
+
"value": 2.771415556231245,
|
| 78 |
+
"lo": 1.43295525583927,
|
| 79 |
+
"hi": 4.109875856623219
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"model": "bridge",
|
| 83 |
+
"feature_set": "hard_no2",
|
| 84 |
+
"value": 2.496059678850094,
|
| 85 |
+
"lo": 0.912988667638017,
|
| 86 |
+
"hi": 4.079130690062171
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"model": "bridge",
|
| 90 |
+
"feature_set": "hard_prices",
|
| 91 |
+
"value": 2.62062591496925,
|
| 92 |
+
"lo": 1.1868895915257238,
|
| 93 |
+
"hi": 4.054362238412777
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"model": "bridge",
|
| 97 |
+
"feature_set": "all",
|
| 98 |
+
"value": 2.4001047258769646,
|
| 99 |
+
"lo": 1.0819277773060407,
|
| 100 |
+
"hi": 3.7182816744478884
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"model": "enet",
|
| 104 |
+
"feature_set": "hard_only",
|
| 105 |
+
"value": 2.622081984570472,
|
| 106 |
+
"lo": 1.2321626043496732,
|
| 107 |
+
"hi": 4.012001364791271
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"model": "enet",
|
| 111 |
+
"feature_set": "hard_trends",
|
| 112 |
+
"value": 2.5062786974179816,
|
| 113 |
+
"lo": 1.1956021229098588,
|
| 114 |
+
"hi": 3.8169552719261044
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"model": "enet",
|
| 118 |
+
"feature_set": "hard_lights",
|
| 119 |
+
"value": 2.622081984570472,
|
| 120 |
+
"lo": 1.2958890611390879,
|
| 121 |
+
"hi": 3.948274908001856
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"model": "enet",
|
| 125 |
+
"feature_set": "hard_no2",
|
| 126 |
+
"value": 2.0694009258166295,
|
| 127 |
+
"lo": 0.6794815455958307,
|
| 128 |
+
"hi": 3.4593203060374282
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"model": "enet",
|
| 132 |
+
"feature_set": "hard_prices",
|
| 133 |
+
"value": 2.6069274801172466,
|
| 134 |
+
"lo": 1.265976164598964,
|
| 135 |
+
"hi": 3.9478787956355292
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"model": "enet",
|
| 139 |
+
"feature_set": "all",
|
| 140 |
+
"value": 2.06940092581663,
|
| 141 |
+
"lo": 0.7471238227565626,
|
| 142 |
+
"hi": 3.3916780288766972
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"model": "midas",
|
| 146 |
+
"feature_set": "hard_only",
|
| 147 |
+
"value": 3.4239859732768743,
|
| 148 |
+
"lo": 1.3072362627081366,
|
| 149 |
+
"hi": 5.540735683845612
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"model": "midas",
|
| 153 |
+
"feature_set": "all",
|
| 154 |
+
"value": 3.2150876254650664,
|
| 155 |
+
"lo": 1.1521296518849318,
|
| 156 |
+
"hi": 5.2780455990452015
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"model": "lgbm",
|
| 160 |
+
"feature_set": "hard_only",
|
| 161 |
+
"value": 2.9411170887132343,
|
| 162 |
+
"lo": 1.3701709870669854,
|
| 163 |
+
"hi": 4.512063190359483
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"model": "lgbm",
|
| 167 |
+
"feature_set": "all",
|
| 168 |
+
"value": 3.101451725661528,
|
| 169 |
+
"lo": 1.5555462561563083,
|
| 170 |
+
"hi": 4.647357195166748
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"model": "dfm",
|
| 174 |
+
"feature_set": "hard_only",
|
| 175 |
+
"value": 3.6750213539314704,
|
| 176 |
+
"lo": 2.013620414584559,
|
| 177 |
+
"hi": 5.336422293278382
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"model": "dfm",
|
| 181 |
+
"feature_set": "all",
|
| 182 |
+
"value": 3.390650768988231,
|
| 183 |
+
"lo": 1.781317465636148,
|
| 184 |
+
"hi": 4.999984072340314
|
| 185 |
+
}
|
| 186 |
+
],
|
| 187 |
+
"origins": [
|
| 188 |
+
{
|
| 189 |
+
"horizon": "M1",
|
| 190 |
+
"origin": "2026-07-31"
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"horizon": "M2",
|
| 194 |
+
"origin": "2026-08-31"
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"horizon": "M3",
|
| 198 |
+
"origin": "2026-09-30"
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"horizon": "latest",
|
| 202 |
+
"origin": "2026-11-05"
|
| 203 |
+
}
|
| 204 |
+
]
|
| 205 |
+
},
|
| 206 |
+
"cpi_yoy": {
|
| 207 |
+
"label": "Headline CPI inflation (y-o-y)",
|
| 208 |
+
"period": "2026-10",
|
| 209 |
+
"release_expected": "2026-11-06",
|
| 210 |
+
"headline": {
|
| 211 |
+
"model": "enet",
|
| 212 |
+
"feature_set": "all",
|
| 213 |
+
"value": 2.3061274905583997,
|
| 214 |
+
"lo": 1.7569674731498859,
|
| 215 |
+
"hi": 2.8552875079669136,
|
| 216 |
+
"rmsfe_full": 0.4361894121007024
|
| 217 |
+
},
|
| 218 |
+
"models": [
|
| 219 |
+
{
|
| 220 |
+
"model": "rw",
|
| 221 |
+
"feature_set": "none",
|
| 222 |
+
"value": 1.986116,
|
| 223 |
+
"lo": 1.2255880000000001,
|
| 224 |
+
"hi": 2.746644
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"model": "ar",
|
| 228 |
+
"feature_set": "none",
|
| 229 |
+
"value": 2.00337733475106,
|
| 230 |
+
"lo": 1.268319552720219,
|
| 231 |
+
"hi": 2.7384351167819014
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"model": "bridge",
|
| 235 |
+
"feature_set": "hard_only",
|
| 236 |
+
"value": 2.0678123266897193,
|
| 237 |
+
"lo": 0.6540992818840159,
|
| 238 |
+
"hi": 3.4815253714954224
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"model": "bridge",
|
| 242 |
+
"feature_set": "hard_trends",
|
| 243 |
+
"value": 1.9620034720025499,
|
| 244 |
+
"lo": 0.4891452396813707,
|
| 245 |
+
"hi": 3.434861704323729
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"model": "bridge",
|
| 249 |
+
"feature_set": "hard_lights",
|
| 250 |
+
"value": 2.0765670156961114,
|
| 251 |
+
"lo": 0.37655284097829655,
|
| 252 |
+
"hi": 3.776581190413926
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"model": "bridge",
|
| 256 |
+
"feature_set": "hard_no2",
|
| 257 |
+
"value": 1.9200449635415748,
|
| 258 |
+
"lo": 0.6043047433915709,
|
| 259 |
+
"hi": 3.2357851836915787
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"model": "bridge",
|
| 263 |
+
"feature_set": "hard_prices",
|
| 264 |
+
"value": 2.2296769839880586,
|
| 265 |
+
"lo": 1.1140155832432992,
|
| 266 |
+
"hi": 3.345338384732818
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"model": "bridge",
|
| 270 |
+
"feature_set": "all",
|
| 271 |
+
"value": 1.975565967816605,
|
| 272 |
+
"lo": 0.6607716191167214,
|
| 273 |
+
"hi": 3.2903603165164887
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"model": "enet",
|
| 277 |
+
"feature_set": "hard_only",
|
| 278 |
+
"value": 1.4688316407918371,
|
| 279 |
+
"lo": 0.7576763015292372,
|
| 280 |
+
"hi": 2.179986980054437
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"model": "enet",
|
| 284 |
+
"feature_set": "hard_trends",
|
| 285 |
+
"value": 1.5680678603109373,
|
| 286 |
+
"lo": 0.8654972705394863,
|
| 287 |
+
"hi": 2.270638450082388
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"model": "enet",
|
| 291 |
+
"feature_set": "hard_lights",
|
| 292 |
+
"value": 1.453471697024606,
|
| 293 |
+
"lo": 0.7403797703699131,
|
| 294 |
+
"hi": 2.166563623679299
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"model": "enet",
|
| 298 |
+
"feature_set": "hard_no2",
|
| 299 |
+
"value": 1.4918525949020902,
|
| 300 |
+
"lo": 0.7879344711507577,
|
| 301 |
+
"hi": 2.1957707186534225
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"model": "enet",
|
| 305 |
+
"feature_set": "hard_prices",
|
| 306 |
+
"value": 2.3240561431880207,
|
| 307 |
+
"lo": 1.7618486314697233,
|
| 308 |
+
"hi": 2.886263654906318
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"model": "enet",
|
| 312 |
+
"feature_set": "all",
|
| 313 |
+
"value": 2.3061274905583997,
|
| 314 |
+
"lo": 1.7569674731498859,
|
| 315 |
+
"hi": 2.8552875079669136
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"model": "lgbm",
|
| 319 |
+
"feature_set": "hard_only",
|
| 320 |
+
"value": 1.7115373151574091,
|
| 321 |
+
"lo": 0.9853831938492812,
|
| 322 |
+
"hi": 2.437691436465537
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"model": "lgbm",
|
| 326 |
+
"feature_set": "hard_trends",
|
| 327 |
+
"value": 1.8052224662652232,
|
| 328 |
+
"lo": 1.0741084133932657,
|
| 329 |
+
"hi": 2.5363365191371807
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"model": "lgbm",
|
| 333 |
+
"feature_set": "hard_lights",
|
| 334 |
+
"value": 1.7343294825051507,
|
| 335 |
+
"lo": 0.9627479740720566,
|
| 336 |
+
"hi": 2.505910990938245
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"model": "lgbm",
|
| 340 |
+
"feature_set": "hard_no2",
|
| 341 |
+
"value": 1.601669895932529,
|
| 342 |
+
"lo": 0.8755157746244011,
|
| 343 |
+
"hi": 2.327824017240657
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"model": "lgbm",
|
| 347 |
+
"feature_set": "hard_prices",
|
| 348 |
+
"value": 2.6627841727515746,
|
| 349 |
+
"lo": 2.042484129772815,
|
| 350 |
+
"hi": 3.283084215730334
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"model": "lgbm",
|
| 354 |
+
"feature_set": "all",
|
| 355 |
+
"value": 2.4313120467587654,
|
| 356 |
+
"lo": 1.8900278903158707,
|
| 357 |
+
"hi": 2.97259620320166
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"model": "dfm",
|
| 361 |
+
"feature_set": "hard_only",
|
| 362 |
+
"value": 1.9848952359169132,
|
| 363 |
+
"lo": 1.2595816683075094,
|
| 364 |
+
"hi": 2.710208803526317
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"model": "dfm",
|
| 368 |
+
"feature_set": "all",
|
| 369 |
+
"value": 1.9386500191541647,
|
| 370 |
+
"lo": 1.1784398057771612,
|
| 371 |
+
"hi": 2.6988602325311684
|
| 372 |
+
}
|
| 373 |
+
],
|
| 374 |
+
"origins": [
|
| 375 |
+
{
|
| 376 |
+
"horizon": "mid",
|
| 377 |
+
"origin": "2026-10-15"
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"horizon": "M0",
|
| 381 |
+
"origin": "2026-10-31"
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"horizon": "PR",
|
| 385 |
+
"origin": "2026-11-05"
|
| 386 |
+
}
|
| 387 |
+
]
|
| 388 |
+
}
|
| 389 |
+
},
|
| 390 |
+
"news": {
|
| 391 |
+
"target": "gdp_yoy",
|
| 392 |
+
"feature_set": "all",
|
| 393 |
+
"from": "2026-09-30",
|
| 394 |
+
"to": "2026-11-05",
|
| 395 |
+
"nowcast_prev": 3.5685982984504783,
|
| 396 |
+
"nowcast_new": 3.3940234709497727,
|
| 397 |
+
"change": -0.17457482750070552,
|
| 398 |
+
"method": "statsmodels news",
|
| 399 |
+
"by_series": [
|
| 400 |
+
{
|
| 401 |
+
"series": "gt_travel",
|
| 402 |
+
"impact": -0.1086785742316695
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"series": "gt_unemploy",
|
| 406 |
+
"impact": -0.04016759006269474
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"series": "vehicle_reg",
|
| 410 |
+
"impact": -0.036649334000346166
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"series": "gt_loan",
|
| 414 |
+
"impact": -0.01628605938699763
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"series": "imports",
|
| 418 |
+
"impact": -0.015136326294564404
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"series": "expressway",
|
| 422 |
+
"impact": -0.009875545780666455
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"series": "electricity",
|
| 426 |
+
"impact": -0.0078903737892733
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"series": "aot_pax",
|
| 430 |
+
"impact": -0.0030941998220819413
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"series": "gt_car_installment",
|
| 434 |
+
"impact": -0.0018230290502531492
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"series": "tourists",
|
| 438 |
+
"impact": -0.00153754350423599
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"series": "gt_prices",
|
| 442 |
+
"impact": -0.000554228192775243
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"series": "agri_price",
|
| 446 |
+
"impact": -0.00011514053254522077
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"series": "no2",
|
| 450 |
+
"impact": 8.707749035095566e-05
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"series": "oil_retail",
|
| 454 |
+
"impact": 0.0003556778514359496
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"series": "promptpay",
|
| 458 |
+
"impact": 0.00046628805263241923
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"series": "pii",
|
| 462 |
+
"impact": 0.0012703540228690478
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"series": "online_prices",
|
| 466 |
+
"impact": 0.0017428370827895715
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"series": "mpi",
|
| 470 |
+
"impact": 0.001969272432995771
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"series": "credit",
|
| 474 |
+
"impact": 0.002661494191363973
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"series": "exports",
|
| 478 |
+
"impact": 0.010002031045984879
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"series": "nightlights",
|
| 482 |
+
"impact": 0.014907920357179036
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"series": "pci",
|
| 486 |
+
"impact": 0.0300623035927379
|
| 487 |
+
}
|
| 488 |
+
]
|
| 489 |
+
}
|
| 490 |
+
}
|
model_specs.json
ADDED
|
@@ -0,0 +1,387 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"package": "thainowcast",
|
| 3 |
+
"version": "0.1.0",
|
| 4 |
+
"code": "https://github.com/thanatchon36/thainowcast",
|
| 5 |
+
"commit": "ce9dd21fafbf9adcd649900c8a006a776bdc6d4c",
|
| 6 |
+
"trained_on": "synthetic test data (seed 20261105); no fitted weights are stored, every model is re-estimated at each origin from the data released by that date",
|
| 7 |
+
"models": {
|
| 8 |
+
"rw": {
|
| 9 |
+
"class": "thainowcast.models.benchmarks.RandomWalk",
|
| 10 |
+
"params": {}
|
| 11 |
+
},
|
| 12 |
+
"ar": {
|
| 13 |
+
"class": "thainowcast.models.benchmarks.AR",
|
| 14 |
+
"params": {
|
| 15 |
+
"max_lags": 4
|
| 16 |
+
}
|
| 17 |
+
},
|
| 18 |
+
"bridge": {
|
| 19 |
+
"class": "thainowcast.models.bridge.Bridge",
|
| 20 |
+
"params": {
|
| 21 |
+
"n_factors": 2
|
| 22 |
+
}
|
| 23 |
+
},
|
| 24 |
+
"midas": {
|
| 25 |
+
"class": "thainowcast.models.midas.MIDAS",
|
| 26 |
+
"params": {
|
| 27 |
+
"n_lags": 12
|
| 28 |
+
}
|
| 29 |
+
},
|
| 30 |
+
"dfm": {
|
| 31 |
+
"class": "thainowcast.models.dfm.DFM",
|
| 32 |
+
"params": {
|
| 33 |
+
"n_factors": 1,
|
| 34 |
+
"factor_order": 1,
|
| 35 |
+
"maxiter": 50,
|
| 36 |
+
"refit": "annual"
|
| 37 |
+
}
|
| 38 |
+
},
|
| 39 |
+
"enet": {
|
| 40 |
+
"class": "thainowcast.models.ml.ElasticNetNowcast",
|
| 41 |
+
"params": {
|
| 42 |
+
"l1_ratios": [
|
| 43 |
+
0.1,
|
| 44 |
+
0.5,
|
| 45 |
+
0.9
|
| 46 |
+
],
|
| 47 |
+
"n_splits": 4
|
| 48 |
+
}
|
| 49 |
+
},
|
| 50 |
+
"lgbm": {
|
| 51 |
+
"class": "thainowcast.models.ml.LightGBMNowcast",
|
| 52 |
+
"params": {
|
| 53 |
+
"n_estimators": 300,
|
| 54 |
+
"learning_rate": 0.03,
|
| 55 |
+
"num_leaves": 4,
|
| 56 |
+
"min_child_samples": 5
|
| 57 |
+
}
|
| 58 |
+
}
|
| 59 |
+
},
|
| 60 |
+
"feature_sets": {
|
| 61 |
+
"hard_only": [
|
| 62 |
+
"mpi",
|
| 63 |
+
"pci",
|
| 64 |
+
"pii",
|
| 65 |
+
"credit",
|
| 66 |
+
"promptpay",
|
| 67 |
+
"exports",
|
| 68 |
+
"imports",
|
| 69 |
+
"tourists",
|
| 70 |
+
"aot_pax",
|
| 71 |
+
"electricity",
|
| 72 |
+
"vehicle_reg",
|
| 73 |
+
"expressway",
|
| 74 |
+
"oil_retail",
|
| 75 |
+
"agri_price"
|
| 76 |
+
],
|
| 77 |
+
"hard_trends": [
|
| 78 |
+
"mpi",
|
| 79 |
+
"pci",
|
| 80 |
+
"pii",
|
| 81 |
+
"credit",
|
| 82 |
+
"promptpay",
|
| 83 |
+
"exports",
|
| 84 |
+
"imports",
|
| 85 |
+
"tourists",
|
| 86 |
+
"aot_pax",
|
| 87 |
+
"electricity",
|
| 88 |
+
"vehicle_reg",
|
| 89 |
+
"expressway",
|
| 90 |
+
"oil_retail",
|
| 91 |
+
"agri_price",
|
| 92 |
+
"gt_unemploy",
|
| 93 |
+
"gt_loan",
|
| 94 |
+
"gt_car_installment",
|
| 95 |
+
"gt_travel",
|
| 96 |
+
"gt_prices"
|
| 97 |
+
],
|
| 98 |
+
"hard_lights": [
|
| 99 |
+
"mpi",
|
| 100 |
+
"pci",
|
| 101 |
+
"pii",
|
| 102 |
+
"credit",
|
| 103 |
+
"promptpay",
|
| 104 |
+
"exports",
|
| 105 |
+
"imports",
|
| 106 |
+
"tourists",
|
| 107 |
+
"aot_pax",
|
| 108 |
+
"electricity",
|
| 109 |
+
"vehicle_reg",
|
| 110 |
+
"expressway",
|
| 111 |
+
"oil_retail",
|
| 112 |
+
"agri_price",
|
| 113 |
+
"nightlights"
|
| 114 |
+
],
|
| 115 |
+
"hard_no2": [
|
| 116 |
+
"mpi",
|
| 117 |
+
"pci",
|
| 118 |
+
"pii",
|
| 119 |
+
"credit",
|
| 120 |
+
"promptpay",
|
| 121 |
+
"exports",
|
| 122 |
+
"imports",
|
| 123 |
+
"tourists",
|
| 124 |
+
"aot_pax",
|
| 125 |
+
"electricity",
|
| 126 |
+
"vehicle_reg",
|
| 127 |
+
"expressway",
|
| 128 |
+
"oil_retail",
|
| 129 |
+
"agri_price",
|
| 130 |
+
"no2"
|
| 131 |
+
],
|
| 132 |
+
"hard_prices": [
|
| 133 |
+
"mpi",
|
| 134 |
+
"pci",
|
| 135 |
+
"pii",
|
| 136 |
+
"credit",
|
| 137 |
+
"promptpay",
|
| 138 |
+
"exports",
|
| 139 |
+
"imports",
|
| 140 |
+
"tourists",
|
| 141 |
+
"aot_pax",
|
| 142 |
+
"electricity",
|
| 143 |
+
"vehicle_reg",
|
| 144 |
+
"expressway",
|
| 145 |
+
"oil_retail",
|
| 146 |
+
"agri_price",
|
| 147 |
+
"online_prices"
|
| 148 |
+
],
|
| 149 |
+
"all": [
|
| 150 |
+
"mpi",
|
| 151 |
+
"pci",
|
| 152 |
+
"pii",
|
| 153 |
+
"credit",
|
| 154 |
+
"promptpay",
|
| 155 |
+
"exports",
|
| 156 |
+
"imports",
|
| 157 |
+
"tourists",
|
| 158 |
+
"aot_pax",
|
| 159 |
+
"electricity",
|
| 160 |
+
"vehicle_reg",
|
| 161 |
+
"expressway",
|
| 162 |
+
"oil_retail",
|
| 163 |
+
"agri_price",
|
| 164 |
+
"gt_unemploy",
|
| 165 |
+
"gt_loan",
|
| 166 |
+
"gt_car_installment",
|
| 167 |
+
"gt_travel",
|
| 168 |
+
"gt_prices",
|
| 169 |
+
"nightlights",
|
| 170 |
+
"no2",
|
| 171 |
+
"online_prices",
|
| 172 |
+
"mobility"
|
| 173 |
+
],
|
| 174 |
+
"none": []
|
| 175 |
+
},
|
| 176 |
+
"backtest": {
|
| 177 |
+
"start": "2015-01",
|
| 178 |
+
"benchmark": "ar",
|
| 179 |
+
"conformal_alpha": 0.2,
|
| 180 |
+
"regimes": {
|
| 181 |
+
"pre_covid": [
|
| 182 |
+
"2015-01",
|
| 183 |
+
"2019-12"
|
| 184 |
+
],
|
| 185 |
+
"covid": [
|
| 186 |
+
"2020-01",
|
| 187 |
+
"2021-12"
|
| 188 |
+
],
|
| 189 |
+
"post_covid": [
|
| 190 |
+
"2022-01",
|
| 191 |
+
"2035-12"
|
| 192 |
+
]
|
| 193 |
+
}
|
| 194 |
+
},
|
| 195 |
+
"targets": {
|
| 196 |
+
"gdp_yoy": {
|
| 197 |
+
"frequency": "Q",
|
| 198 |
+
"origins": [
|
| 199 |
+
"M1",
|
| 200 |
+
"M2",
|
| 201 |
+
"M3",
|
| 202 |
+
"PR"
|
| 203 |
+
],
|
| 204 |
+
"final_origin": "PR",
|
| 205 |
+
"grid": [
|
| 206 |
+
{
|
| 207 |
+
"model": "rw",
|
| 208 |
+
"feature_set": "none"
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"model": "ar",
|
| 212 |
+
"feature_set": "none"
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"model": "bridge",
|
| 216 |
+
"feature_set": "hard_only"
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"model": "bridge",
|
| 220 |
+
"feature_set": "hard_trends"
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"model": "bridge",
|
| 224 |
+
"feature_set": "hard_lights"
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"model": "bridge",
|
| 228 |
+
"feature_set": "hard_no2"
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"model": "bridge",
|
| 232 |
+
"feature_set": "hard_prices"
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"model": "bridge",
|
| 236 |
+
"feature_set": "all"
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"model": "enet",
|
| 240 |
+
"feature_set": "hard_only"
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"model": "enet",
|
| 244 |
+
"feature_set": "hard_trends"
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"model": "enet",
|
| 248 |
+
"feature_set": "hard_lights"
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"model": "enet",
|
| 252 |
+
"feature_set": "hard_no2"
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"model": "enet",
|
| 256 |
+
"feature_set": "hard_prices"
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"model": "enet",
|
| 260 |
+
"feature_set": "all"
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"model": "midas",
|
| 264 |
+
"feature_set": "hard_only"
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"model": "midas",
|
| 268 |
+
"feature_set": "all"
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"model": "lgbm",
|
| 272 |
+
"feature_set": "hard_only"
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"model": "lgbm",
|
| 276 |
+
"feature_set": "all"
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"model": "dfm",
|
| 280 |
+
"feature_set": "hard_only"
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"model": "dfm",
|
| 284 |
+
"feature_set": "all"
|
| 285 |
+
}
|
| 286 |
+
]
|
| 287 |
+
},
|
| 288 |
+
"cpi_yoy": {
|
| 289 |
+
"frequency": "M",
|
| 290 |
+
"origins": [
|
| 291 |
+
"M0",
|
| 292 |
+
"PR"
|
| 293 |
+
],
|
| 294 |
+
"final_origin": "PR",
|
| 295 |
+
"grid": [
|
| 296 |
+
{
|
| 297 |
+
"model": "rw",
|
| 298 |
+
"feature_set": "none"
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"model": "ar",
|
| 302 |
+
"feature_set": "none"
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"model": "bridge",
|
| 306 |
+
"feature_set": "hard_only"
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"model": "bridge",
|
| 310 |
+
"feature_set": "hard_trends"
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"model": "bridge",
|
| 314 |
+
"feature_set": "hard_lights"
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"model": "bridge",
|
| 318 |
+
"feature_set": "hard_no2"
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"model": "bridge",
|
| 322 |
+
"feature_set": "hard_prices"
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"model": "bridge",
|
| 326 |
+
"feature_set": "all"
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"model": "enet",
|
| 330 |
+
"feature_set": "hard_only"
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"model": "enet",
|
| 334 |
+
"feature_set": "hard_trends"
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"model": "enet",
|
| 338 |
+
"feature_set": "hard_lights"
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"model": "enet",
|
| 342 |
+
"feature_set": "hard_no2"
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"model": "enet",
|
| 346 |
+
"feature_set": "hard_prices"
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"model": "enet",
|
| 350 |
+
"feature_set": "all"
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"model": "lgbm",
|
| 354 |
+
"feature_set": "hard_only"
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"model": "lgbm",
|
| 358 |
+
"feature_set": "hard_trends"
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"model": "lgbm",
|
| 362 |
+
"feature_set": "hard_lights"
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"model": "lgbm",
|
| 366 |
+
"feature_set": "hard_no2"
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"model": "lgbm",
|
| 370 |
+
"feature_set": "hard_prices"
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"model": "lgbm",
|
| 374 |
+
"feature_set": "all"
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"model": "dfm",
|
| 378 |
+
"feature_set": "hard_only"
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"model": "dfm",
|
| 382 |
+
"feature_set": "all"
|
| 383 |
+
}
|
| 384 |
+
]
|
| 385 |
+
}
|
| 386 |
+
}
|
| 387 |
+
}
|