wsim-vis-python.qmd

---
title: "Untitled"
format: html
jupyter: python3
---
## Read in WSIM-GLDAS

```{python}
import xarray as xr

file_path = "composite_12mo.nc"
wsim_gldas = xr.open_dataset(file_path, engine = 'h5netcdf')
```

## Temporal/Variable Subsetting
```{python}
import pandas as pd
wsim_gldas

keeps = pd.date_range(start="2000-12-01", end="2014-12-01", freq = "YS-DEC")
keeps

wsim_gldas = wsim_gldas.sel(time= keeps)
wsim_gldas = wsim_gldas[["deficit", "crs"]]
wsim_gldas
```

```{python}
p = wsim_gldas.deficit.plot(x="lon", y="lat", col="time", col_wrap = 3, cmap = "Reds_r", aspect = 2)
```

## Read in geoBoundaries

```{python}
import requests
import geopandas as gp

usa = requests.get("https://www.geoboundaries.org/api/current/gbOpen/USA/ADM1/")

usa = usa.json()
usa['gjDownloadURL']
usa = gp.read_file(usa['gjDownloadURL'])

drops = ["Alaska", "Hawaii", "American Samoa", "Puerto Rico", "Commonwealth of the Northern Mariana Islands", "Guam", "United States Virgin Islands"]

usa = usa[~usa.shapeName.isin(drops)]
usa.boundary.plot()
```

```{python}
import rioxarray as rio
wsim_gldas = wsim_gldas.rio.write_crs("epsg: 4326")
wsim_gldas = wsim_gldas.rio.clip(usa.geometry.values)
```

```{python fig.width="100%"}
wsim_gldas.deficit.plot(x="lon", y="lat", col="time", col_wrap = 3, cmap = "Reds_r", aspect = 1.75, size =1.25)
```

## Extract the California Counties
```{python}
usa_counties = requests.get("https://www.geoboundaries.org/api/current/gbOpen/USA/ADM2/")

usa_counties = usa_counties.json()
usa_counties['gjDownloadURL']
usa_counties = gp.read_file(usa_counties['gjDownloadURL'])

california = usa[usa["shapeName"].str.contains("California")]
california_counties = usa_counties.overlay(california, how='intersection')
california_counties.plot()
```

## WSIM-GLDAS 1-Month Integration
```{python}
file_path = "composite_1mo.nc"
wsim_gldas_1mo = xr.open_dataset(file_path, engine = 'h5netcdf')
```

## Temporal/Variable Subsetting
```{python}
keeps = pd.date_range(start="2014-01-01", end="2014-12-01", freq = "MS")

wsim_gldas_1mo = wsim_gldas_1mo.sel(time= keeps)
wsim_gldas_1mo = wsim_gldas_1mo[["deficit", "crs"]]
wsim_gldas_1mo
```

```{python}
wsim_gldas_1mo = wsim_gldas_1mo.rio.write_crs("epsg: 4326")
wsim_gldas_1mo = wsim_gldas_1mo.rio.clip(california_counties.geometry.values)
```

```{python}
wsim_gldas_1mo.deficit.plot(x="lon", y="lat", col="time", col_wrap = 3, cmap = "Reds_r", aspect = 1, size =2.5)
```

## Choropleths
```{python}
from exactextract import exact_extract
import calendar

cc_summaries = exact_extract(wsim_gldas_1mo.deficit, california_counties, "mean", output = 'pandas', include_cols = "shapeName_1", include_geom = True)

col_names = [["county"], calendar.month_name[1:], ["geometry"]]
col_names = sum(col_names, [])

cc_summaries.columns = col_names

cc_summaries.plot("January", cmap = "Reds_r")
```


## Re-Classify the WSIM-GLDAS Raster

```{python}
import numpy

wsim_bins = [numpy.inf, 0, -3, -5, -10, -20, -40, -60, numpy.NINF]

wsim_class = xr.apply_ufunc(
    numpy.digitize,
    wsim_gldas_1mo,
    wsim_bins)

# wsim_class.where(wsim_class['deficit'] == -9223372036854775808) = 7

```

# GPW

```{python}
file_path = "gpw_v4_population_count_rev11_2015_15_min.tif" 
# Open with rioxarray
gpw = rio.open_rasterio(file_path)
```

```{python}
cal_wsim_gpw = exact_extract(
    wsim_class.deficit, 
    california_counties, 
    ['coverage', 'values', 'weights'], 
    output = 'pandas', 
    include_geom = False, 
    weights = gpw)
```

```{python}
explode_cols = list(cal_wsim_gpw.columns)
cal_wsim_gpw = cal_wsim_gpw.explode(explode_cols)
```

```{python}
cal_wsim_cov = cal_wsim_gpw.filter(like = 'coverage', axis = 1)
cal_wsim_cov.columns = calendar.month_name[1:]
cal_wsim_cov= pd.melt(cal_wsim_cov, var_name='month', value_name= 'coverage')
```

```{python}
cal_wsim_val = cal_wsim_gpw.filter(like = 'values', axis = 1)
cal_wsim_val.columns = calendar.month_name[1:]
cal_wsim_val= pd.melt(cal_wsim_val, var_name='month', value_name = 'wsim_class')
```

```{python}
cal_wsim_weight = cal_wsim_gpw.filter(like = 'weight', axis = 1)
cal_wsim_weight.columns = calendar.month_name[1:]
cal_wsim_weight= pd.melt(cal_wsim_weight, var_name='month', value_name = 'cell_pop_count')
```

```{python}
test = pd.concat(
    [cal_wsim_cov, 
    cal_wsim_val["wsim_class"],
    cal_wsim_weight["cell_pop_count"]], 
    axis=1)
```

```{python}
test['wsim_class'].replace(-9223372036854775808, 7)
```

```{python}
test["wsim_class_pop"] = test["coverage"]*test["cell_pop_count"]
test.wsim_class_pop = test['wsim_class_pop'].astype('float').round(0)
test = test.groupby(['month', 'wsim_class'])['wsim_class_pop'].sum().reset_index()
test['month_pop'] = test.groupby(['month'])['wsim_class_pop'].transform('sum')
test['wsim_class_frac'] = test['wsim_class_pop'] / test['month_pop']
test
```

```{python}
test['wsim_class'] = test['wsim_class'].astype('category')
test['wsim_class'] = test['wsim_class'].cat.rename_categories(
    {0: "0", 1: "-3", 2: "-5", 3: "-10", 4: "-20", 5: "-40", 6: "-50", 7: "-60"})

```
```{python}
test["month"] = pd.Categorical(test["month"],
                             categories=["January", "February", "March", "April", "May", "June", "July",
                                         "August", "September", "October", "November", "December"],
                             ordered=True)
leg_colors=['#9B0039',
    # -50 to -40
    '#D44135',
    # -40 to -20
    '#FF8D43',
    # -20 to -10
    '#FFC754',
    # -10 to -5
    '#FFEDA3',
    # -5 to -3
    '#fffdc7',
    # -3 to 0
    '#FFF4C7',
    # 0-3
    "#FFFFFF"]
```

```{python}
from plotnine import *

(ggplot(test, aes('month', 'wsim_class_frac', fill = 'wsim_class', group='wsim_class'))+ 
scale_fill_manual(values = leg_colors[::-1])+
geom_bar(stat='identity', position='stack')+ 
 labs(title = "Monthly Fraction of Population Under Water Deficits in California During 2014",
                subtitle = "Categorized by Intensity of Deficit Return Period",
                x = "",
                y = "Fraction of Population*",
                caption = "*Population derived from Gridded Population of the World (2015)",
                fill = "Deficit Class")+
theme_minimal()+
theme(axis_text_x=element_text(rotation=25, hjust=1))
)
```