Spatial Causal Inference

class: center, middle, inverse, title-slide

.title[
# Spatial Causal Inference
]
.subtitle[
## spcausal : A Framework for Spatial Causal Inference
]
.author[
### Wenbo Lv
]
.date[
### 2024-10-18
]

---

class: inverse, middle, center

body {
  font-family: 'Roboto Condensed', 'Avenir Next', 'Helvetica Neue', 'Helvetica', sans-serif;
  font-weight: 300;
}
h1, h2, h3 {
  font-family: 'Roboto', 'Avenir Next', 'Helvetica Neue', 'Helvetica', sans-serif;
  font-weight: 600;
}
.remark-code, .remark-inline-code { 
font-family: 'Inconsolata', 'Consolas', Monaco, monospace; 
}

.title-slide {
 background-image: url("https://spatlyu.github.io/materials/figures/ai4city_ppt_title_bg1.png");
 background-position: center;
 background-size: contain;
}
</style>

# Causal Inference Methods In Geography

.footnote[The following is a Discipline-based perspective]

---

## 1. Based on Spatial Statistics

- **Spatial Dependence** & **Spatial Association**

- **Spatial Heterogeneity**

- **Spatial Similarity**

[On spatial effects in geographical analysis](https://www.geog.com.cn/EN/10.11821/dlxb202303001)

[Geodetector: Principle and prospective](https://doi.org/10.11821/dlxb201701010)

[Modeling of spatial stratified heterogeneity](https://doi.org/10.1080/15481603.2022.2126375)

[Statistical Modeling of Spatially Stratified Heterogeneous Data](https://doi.org/10.1080/24694452.2023.2289982)

[Causal inference in spatial statistics](https://www.sciencedirect.com/science/article/pii/S2211675322000173)

[Spatial Causality: A Systematic Review on Spatial Causal Inference](https://onlinelibrary.wiley.com/doi/abs/10.1111/gean.12312)

---

## 2. Based on Spatial Econometrics

- **Regression Discontinuity Design**

- **Difference In Difference**

[Conduct Multiple Types of Geographic Regression Discontinuity Designs(SpatialRDD)](https://axlehner.github.io/SpatialRDD/index.html)

[Spatial Difference-in-differences (SpatialDID)](https://www.sciencedirect.com/science/article/pii/S0165176515004371)

[Evidence on the impact of sustained exposure to air pollution on life expectancy from China’s Huai River policy](https://www.pnas.org/doi/full/10.1073/pnas.1300018110)

---

## 3. Based on Spatial Epidemiology

- **Spatial Causal Mediation Analysis**

- **Agent-Based Modeling**

- **Bayesian Spatial Models**

- **Spatial Survival Analysis**

- **Spatial Propensity Score Matching**

[Bayesian inference with INLA](https://becarioprecario.bitbucket.io/inla-gitbook/)

[Dynamic Time Series Models using R-INLA: An Applied Perspective](https://ramanbala.github.io/dynamic-time-series-models-R-INLA/)

[Advanced Spatial Modeling with Stochastic Partial Differential Equations Using R and INLA ](https://becarioprecario.bitbucket.io/spde-gitbook/)

[Geospatial Health Data: Modeling and Visualization with R-INLA and Shiny](https://www.paulamoraga.com/book-geospatial/)

---

## 4. Based on Causal Machine Learning

- **Causal Graph Learning**

- **Causal Forest**

and so on (I don't know much about it)

[**DoWhy** documentation](https://www.pywhy.org/dowhy)

[**CausalML** documentation](https://causalml.readthedocs.io/en/latest/about.html)

---

## 5.  Based on Dynamic  Systems  Modeling

- **Temporal State space reconstruction**

- **Spatial State space reconstruction**

[Detecting Causality in Complex Ecosystems](https://www.science.org/doi/10.1126/science.1227079)

[Partial cross mapping eliminates indirect causal influences](https://www.nature.com/articles/s41467-020-16238-0)

[Detecting Causality from Nonlinear Dynamics with Short-term Time Series](https://www.nature.com/articles/srep07464)

[Inferring causation from time series in Earth system sciences](https://www.nature.com/articles/s41467-019-10105-3)

[Spatial convergent cross mapping to detect causal relationships from short time series](https://esajournals.onlinelibrary.wiley.com/doi/full/10.1890/14-1479.1)

[Causal inference from cross-sectional earth system data with geographical convergent cross mapping](https://www.nature.com/articles/s41467-023-41619-6)

---
class: inverse, middle, center

# Spatial Causal Inference Ecosystem in R language

---

### Related Ecosystem

Data Clean -> tidyverse

Statistics Inference -> tidymodel & easystats

Bayesian -> INLA & Stan & tidybayes & R2WinBUGS

Machine Learning -> mlr3verse & h2o

Deep Learning -> torch & tensorflow & keras & mxnet

Geospatial Processing ->  sf & terra & stars & whitebox & qgisprocess

Spatial Statistics -> spdep & spatstat & spatialreg & gstat & GD & GWmodel

### Causal Inference In R

[`CRAN Views : Causal Inference`](https://cran.r-project.org/web/views/CausalInference.html)

### So, Where is the unified framework of spatial causal inference in R?

---
class: inverse, middle, center

# spcausal : A Framework for Spatial Causal Inference

---

## Problems to be solved

[Two Language Problem](https://scientificcoder.com/how-to-solve-the-two-language-problem)

Complex interdisciplinary and cross-programming language background

Diverse and robust algorithm integration

User-friendly API

Comprehensive reference documentation

Sustained community vitality

---

## What will we do?

- Testing and development of the original algorithm in R version

- Further optimization of the computation process

- Expanding to other programming languages

<div class="plotly html-widget html-fill-item" id="htmlwidget-7d255079791f89b65f64" style="width:100%;height:400px;"></div>
<script type="application/json" data-for="htmlwidget-7d255079791f89b65f64">{"x":{"visdat":{"5e7c5e8c5a64":["function () ","plotlyVisDat"]},"cur_data":"5e7c5e8c5a64","attrs":{"5e7c5e8c5a64":{"mode":"lines","alpha_stroke":1,"sizes":[10,100],"spans":[1,20],"type":"scatter"},"5e7c5e8c5a64.1":{"mode":"lines","alpha_stroke":1,"sizes":[10,100],"spans":[1,20],"type":"scatter","x":["2024-03-05 00:00:00.000000","2025-12-24 00:00:00.000000"],"y":3,"line":{"color":"#8DD3C7","width":100},"showlegend":false,"hoverinfo":"text","text":"Step 1<\/b> from 2024-03-05<\/b> to 2025-12-24<\/b>","inherit":true},"5e7c5e8c5a64.2":{"mode":"text","alpha_stroke":1,"sizes":[10,100],"spans":[1,20],"type":"scatter","x":"2025-01-28 12:00:00.000000","y":3,"text":"Step 1","textfont":{"family":"Arial","size":14,"color":"rgba(0,0,0,1)"},"textposition":"center","showlegend":false,"hoverinfo":"none","inherit":true},"5e7c5e8c5a64.3":{"mode":"lines","alpha_stroke":1,"sizes":[10,100],"spans":[1,20],"type":"scatter","x":["2025-12-25 00:00:00.000000","2030-12-25 00:00:00.000000"],"y":2,"line":{"color":"#FFFFB3","width":100},"showlegend":false,"hoverinfo":"text","text":"Step 2<\/b> from 2025-12-25<\/b> to 2030-12-25<\/b>","inherit":true},"5e7c5e8c5a64.4":{"mode":"text","alpha_stroke":1,"sizes":[10,100],"spans":[1,20],"type":"scatter","x":"2028-06-25 00:00:00.000000","y":2,"text":"Step 2","textfont":{"family":"Arial","size":14,"color":"rgba(0,0,0,1)"},"textposition":"center","showlegend":false,"hoverinfo":"none","inherit":true},"5e7c5e8c5a64.5":{"mode":"lines","alpha_stroke":1,"sizes":[10,100],"spans":[1,20],"type":"scatter","x":["2030-12-30 00:00:00.000000","2035-01-01 00:00:00.000000"],"y":1,"line":{"color":"#BEBADA","width":100},"showlegend":false,"hoverinfo":"text","text":"Step 3<\/b> from 2030-12-30<\/b> to 2035-01-01<\/b>","inherit":true},"5e7c5e8c5a64.6":{"mode":"text","alpha_stroke":1,"sizes":[10,100],"spans":[1,20],"type":"scatter","x":"2032-12-30 12:00:00.000000","y":1,"text":"Step 3","textfont":{"family":"Arial","size":14,"color":"rgba(0,0,0,1)"},"textposition":"center","showlegend":false,"hoverinfo":"none","inherit":true}},"layout":{"margin":{"b":40,"l":60,"t":25,"r":10},"hovermode":"closest","plot_bgcolor":"#FCFCFC","xaxis":{"domain":[0,1],"automargin":true,"linewidth":1,"mirror":true,"showgrid":true,"gridcolor":"grey90","title":""},"yaxis":{"domain":[0,1],"automargin":true,"linewidth":1,"mirror":true,"range":[0,4],"showgrid":false,"title":"","tickmode":"array","tickvals":[2],"ticktext":[""]},"showlegend":false},"source":"A","config":{"modeBarButtonsToAdd":["hoverclosest","hovercompare"],"showSendToCloud":false},"data":[{"mode":"lines","type":"scatter","marker":{"color":"rgba(31,119,180,1)","line":{"color":"rgba(31,119,180,1)"}},"error_y":{"color":"rgba(31,119,180,1)"},"error_x":{"color":"rgba(31,119,180,1)"},"line":{"color":"rgba(31,119,180,1)"},"xaxis":"x","yaxis":"y","frame":null},{"mode":"lines","type":"scatter","x":["2024-03-05 00:00:00.000000","2025-12-24 00:00:00.000000"],"y":[3,3],"line":{"color":"#8DD3C7","width":100},"showlegend":false,"hoverinfo":["text","text"],"text":["Step 1<\/b> from 2024-03-05<\/b> to 2025-12-24<\/b>","Step 1<\/b> from 2024-03-05<\/b> to 2025-12-24<\/b>"],"marker":{"color":"rgba(255,127,14,1)","line":{"color":"rgba(255,127,14,1)"}},"error_y":{"color":"rgba(255,127,14,1)"},"error_x":{"color":"rgba(255,127,14,1)"},"xaxis":"x","yaxis":"y","frame":null},{"mode":"text","type":"scatter","x":["2025-01-28 12:00:00.000000"],"y":[3],"text":"Step 1","textfont":{"family":"Arial","size":14,"color":"rgba(0,0,0,1)"},"textposition":"center","showlegend":false,"hoverinfo":"none","marker":{"color":"rgba(44,160,44,1)","line":{"color":"rgba(44,160,44,1)"}},"error_y":{"color":"rgba(44,160,44,1)"},"error_x":{"color":"rgba(44,160,44,1)"},"line":{"color":"rgba(44,160,44,1)"},"xaxis":"x","yaxis":"y","frame":null},{"mode":"lines","type":"scatter","x":["2025-12-25 00:00:00.000000","2030-12-25 00:00:00.000000"],"y":[2,2],"line":{"color":"#FFFFB3","width":100},"showlegend":false,"hoverinfo":["text","text"],"text":["Step 2<\/b> from 2025-12-25<\/b> to 2030-12-25<\/b>","Step 2<\/b> from 2025-12-25<\/b> to 2030-12-25<\/b>"],"marker":{"color":"rgba(214,39,40,1)","line":{"color":"rgba(214,39,40,1)"}},"error_y":{"color":"rgba(214,39,40,1)"},"error_x":{"color":"rgba(214,39,40,1)"},"xaxis":"x","yaxis":"y","frame":null},{"mode":"text","type":"scatter","x":["2028-06-25 00:00:00.000000"],"y":[2],"text":"Step 2","textfont":{"family":"Arial","size":14,"color":"rgba(0,0,0,1)"},"textposition":"center","showlegend":false,"hoverinfo":"none","marker":{"color":"rgba(148,103,189,1)","line":{"color":"rgba(148,103,189,1)"}},"error_y":{"color":"rgba(148,103,189,1)"},"error_x":{"color":"rgba(148,103,189,1)"},"line":{"color":"rgba(148,103,189,1)"},"xaxis":"x","yaxis":"y","frame":null},{"mode":"lines","type":"scatter","x":["2030-12-30 00:00:00.000000","2035-01-01 00:00:00.000000"],"y":[1,1],"line":{"color":"#BEBADA","width":100},"showlegend":false,"hoverinfo":["text","text"],"text":["Step 3<\/b> from 2030-12-30<\/b> to 2035-01-01<\/b>","Step 3<\/b> from 2030-12-30<\/b> to 2035-01-01<\/b>"],"marker":{"color":"rgba(140,86,75,1)","line":{"color":"rgba(140,86,75,1)"}},"error_y":{"color":"rgba(140,86,75,1)"},"error_x":{"color":"rgba(140,86,75,1)"},"xaxis":"x","yaxis":"y","frame":null},{"mode":"text","type":"scatter","x":["2032-12-30 12:00:00.000000"],"y":[1],"text":"Step 3","textfont":{"family":"Arial","size":14,"color":"rgba(0,0,0,1)"},"textposition":"center","showlegend":false,"hoverinfo":"none","marker":{"color":"rgba(227,119,194,1)","line":{"color":"rgba(227,119,194,1)"}},"error_y":{"color":"rgba(227,119,194,1)"},"error_x":{"color":"rgba(227,119,194,1)"},"line":{"color":"rgba(227,119,194,1)"},"xaxis":"x","yaxis":"y","frame":null}],"highlight":{"on":"plotly_click","persistent":false,"dynamic":false,"selectize":false,"opacityDim":0.20000000000000001,"selected":{"opacity":1},"debounce":0},"shinyEvents":["plotly_hover","plotly_click","plotly_selected","plotly_relayout","plotly_brushed","plotly_brushing","plotly_clickannotation","plotly_doubleclick","plotly_deselect","plotly_afterplot","plotly_sunburstclick"],"base_url":"https://plot.ly"},"evals":[],"jsHooks":[]}</script>

---
class: center, middle
background-image: url("https://spatlyu.github.io/materials/figures/ai4city_ppt_title_bg4.png")
background-position: center
background-size: contain

# Thanks

### Wenbo Lv

### lyu.geosocial@gmail.com