File size: 12,943 Bytes
a5f11f1
9f2e9b8
a5f11f1
1f9c86b
 
ceee7a4
 
6001575
a5f11f1
ceee7a4
 
a5f11f1
ceee7a4
 
7d61895
 
 
1f9c86b
 
 
ceee7a4
1f9c86b
 
 
 
ceee7a4
 
 
5db2c30
ceee7a4
7d61895
 
 
ceee7a4
 
 
 
 
 
 
 
 
 
 
 
 
 
5db2c30
a5f11f1
7d61895
 
 
a5f11f1
 
 
 
 
 
 
 
 
7d61895
1f9c86b
7d61895
1f9c86b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
55c0b39
 
 
 
 
1f9c86b
55c0b39
 
1f9c86b
55c0b39
 
 
 
 
 
 
 
 
 
 
 
 
1f9c86b
 
 
55c0b39
1f9c86b
55c0b39
 
 
 
 
 
 
 
 
 
7d61895
 
 
 
 
 
 
 
 
 
 
 
55c0b39
 
 
 
 
7d61895
 
 
 
55c0b39
7d61895
 
 
1f9c86b
7d61895
 
1f9c86b
55c0b39
7d61895
 
 
 
 
 
 
55c0b39
 
 
 
 
 
 
 
7d61895
 
 
 
55c0b39
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7d61895
 
55c0b39
 
 
 
 
1f9c86b
55c0b39
 
 
 
 
 
 
7d61895
 
 
55c0b39
 
 
 
 
 
 
7d61895
 
55c0b39
 
 
 
 
 
 
 
 
 
 
 
 
 
7d61895
55c0b39
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7d61895
55c0b39
7d61895
 
 
 
 
 
 
55c0b39
 
 
 
7d61895
55c0b39
 
 
7d61895
55c0b39
 
 
 
 
 
 
 
 
 
7d61895
 
 
1f9c86b
7d61895
55c0b39
 
 
 
 
 
 
 
 
 
 
 
 
 
1f9c86b
55c0b39
 
1f9c86b
7d61895
 
 
 
 
55c0b39
 
 
 
 
 
 
 
 
 
 
 
 
 
7d61895
55c0b39
7d61895
 
55c0b39
 
 
 
7d61895
55c0b39
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
# app.py
import os
import io
import time
import requests
import pandas as pd
import numpy as np
import geopandas as gpd
import matplotlib.pyplot as plt
from shapely.geometry import Point
import folium
import gradio as gr
from PIL import Image

# ----------------------------
# 設定
# ----------------------------
GSI_USER_AGENT = os.environ.get(
    "GSI_USER_AGENT",
    "jp-gsi-geocoding-demo (contact: your_email@example.com)"  # 連絡先付き推奨
)
GSI_TIMEOUT_SEC = float(os.environ.get("GSI_TIMEOUT_SEC", "10"))
GEOCODE_DELAY_SEC = float(os.environ.get("GSI_RATE_LIMIT_SEC", "0.5"))  # マナーとして少し待機
GSI_GEOCODE_URL = "https://msearch.gsi.go.jp/address-search/AddressSearch"

CACHE_DIR = "data/cache"
os.makedirs(CACHE_DIR, exist_ok=True)
CACHE_PATH = os.path.join(CACHE_DIR, "geocode_cache.csv")
DEFAULT_ZIP = "data/japan_ver85.zip"

# ----------------------------
# キャッシュ
# ----------------------------
def load_cache():
    if os.path.exists(CACHE_PATH):
        try:
            df = pd.read_csv(CACHE_PATH)
            if set(["address_input", "lat", "lon", "CF"]).issubset(df.columns):
                return df
        except Exception:
            pass
    return pd.DataFrame(columns=["address_input", "lat", "lon", "CF"])

def save_cache(df_cache):
    try:
        df_cache.to_csv(CACHE_PATH, index=False)
    except Exception:
        pass

# ----------------------------
# Shapefile 読み込み
# ----------------------------
def load_gdf_from_zip(zip_path: str) -> gpd.GeoDataFrame:
    gdf = gpd.read_file(f"zip://{zip_path}")  # , engine="pyogrio"
    try:
        if gdf.crs:
            gdf = gdf.to_crs("EPSG:4326")
    except Exception:
        pass
    return gdf

# ----------------------------
# 国土地理院 ジオコーダ
# ----------------------------
def make_gsi_session() -> requests.Session:
    s = requests.Session()
    s.headers.update({"User-Agent": GSI_USER_AGENT})
    return s

def gsi_geocode_once(address: str, session: requests.Session) -> tuple[float, float]:
    """
    国土地理院 住所検索APIを1回呼び出し、(lat, lon) を返す。失敗時は (nan, nan)。
    返却座標は [lon, lat] なので順を入れ替えて返す。
    """
    try:
        # 空やnan文字列はスキップ
        if not address or address.strip() == "" or address.strip().lower() in ("nan", "none"):
            return (np.nan, np.nan)

        resp = session.get(GSI_GEOCODE_URL, params={"q": address}, timeout=GSI_TIMEOUT_SEC)
        if not resp.ok:
            return (np.nan, np.nan)
        data = resp.json()
        # 返り値は配列(候補リスト)。最上位候補を採用
        if isinstance(data, list) and len(data) > 0:
            feat = data[0]
            coords = (feat.get("geometry") or {}).get("coordinates") or []
            if isinstance(coords, (list, tuple)) and len(coords) >= 2:
                lon, lat = coords[0], coords[1]
                # 数値化チェック
                lat = float(lat)
                lon = float(lon)
                return (lat, lon)
    except Exception:
        pass
    return (np.nan, np.nan)

def geocode_with_cache(addresses, CFs, use_internet=True):
    cache = load_cache()
    cache_map = {row["address_input"]: (row["lat"], row["lon"], row["CF"]) for _, row in cache.iterrows()}
    results = []
    session = make_gsi_session() if use_internet else None

    for a, cf in zip(addresses, CFs):
        a = "" if (a is None or (isinstance(a, float) and np.isnan(a))) else str(a).strip()
        cf = "" if (cf is None or (isinstance(cf, float) and np.isnan(cf))) else str(cf)

        # cache hit
        if a in cache_map:
            lat, lon, _cached_cf = cache_map[a]
            if pd.notna(lat) and pd.notna(lon):
                results.append({"address_input": a, "CF": cf, "lat": lat, "lon": lon})
                continue

        if not use_internet:
            results.append({"address_input": a, "CF": cf, "lat": np.nan, "lon": np.nan})
            continue

        lat, lon = gsi_geocode_once(a, session)
        # マナーとして小休止
        time.sleep(GEOCODE_DELAY_SEC)

        # キャッシュ更新
        cache = cache[cache["address_input"] != a]
        cache = pd.concat(
            [cache, pd.DataFrame([{"address_input": a, "lat": lat, "lon": lon, "CF": cf}])],
            ignore_index=True
        )
        save_cache(cache)
        results.append({"address_input": a, "CF": cf, "lat": lat, "lon": lon})

    return pd.DataFrame(results)

# ----------------------------
# 可視化(matplotlib)
# ----------------------------
def plot_map_png(
    gdf_pref: gpd.GeoDataFrame,
    gdf_pts: gpd.GeoDataFrame,
    line_width: float = 0.6,
    marker_size: int = 24,
    legend_shrink: float = 0.6,
    legend_fontsize: int = 8,
    figsize=(7, 7),
) -> Image.Image:
    fig, ax = plt.subplots(figsize=figsize)
    gdf_pref.boundary.plot(ax=ax, linewidth=line_width, color="black")

    gdf_pts_valid = gdf_pts[gdf_pts.geometry.notnull()]
    if not gdf_pts_valid.empty:
        cf_num = pd.to_numeric(
            gdf_pts_valid.get("CF", pd.Series([np.nan]*len(gdf_pts_valid))),
            errors="coerce"
        )
        gdf_pts_valid.assign(CF_num=cf_num).plot(
            ax=ax,
            column="CF_num",
            cmap="OrRd",
            markersize=max(2, int(marker_size)),
            alpha=0.85,
            legend=True,
            legend_kwds={"shrink": legend_shrink},
        )
        try:
            for _ax in fig.axes:
                if _ax is not ax:
                    _ax.tick_params(labelsize=legend_fontsize)
        except Exception:
            pass

    ax.set_axis_off()
    plt.tight_layout()
    buf = io.BytesIO()
    fig.savefig(buf, format="png", dpi=200)
    plt.close(fig)
    buf.seek(0)
    return Image.open(buf)

# ----------------------------
# 可視化(folium)
# ----------------------------
def make_folium_html(gdf_pref: gpd.GeoDataFrame, gdf_pts: gpd.GeoDataFrame, marker_size: int = 24):
    gdf_pts_valid = gdf_pts[gdf_pts.geometry.notnull()]
    if not gdf_pts_valid.empty:
        center_lat = gdf_pts_valid.geometry.y.median()
        center_lon = gdf_pts_valid.geometry.x.median()
        zoom = 6
    else:
        center_lat, center_lon, zoom = 35.6812, 139.7671, 5

    m = folium.Map(location=[center_lat, center_lon], zoom_start=zoom)

    try:
        folium.GeoJson(gdf_pref.to_json(), name="prefecture").add_to(m)
    except Exception:
        pass

    circle_radius = max(3, int(marker_size // 3))

    for _, r in gdf_pts_valid.iterrows():
        lat, lon = r.geometry.y, r.geometry.x
        popup = f"{r.get('address_input','(no addr)')}<br>CF:{r.get('CF','')}"
        folium.CircleMarker(
            location=(float(lat), float(lon)),
            radius=circle_radius,
            fill=True,
            fill_opacity=0.9,
            popup=popup,
        ).add_to(m)

    return m._repr_html_()

# ----------------------------
# 実行パイプライン
# ----------------------------
def _parse_indexer(x):
    try:
        return int(x)
    except Exception:
        return x

def run(zip_file, excel_file, sheet_name, header_row, address_col, power_col,
        use_inet, line_width, marker_size, legend_shrink, legend_fontsize):
    # Shapefile
    if zip_file is not None and hasattr(zip_file, "name") and os.path.exists(zip_file.name):
        zip_path = zip_file.name
    elif os.path.exists(DEFAULT_ZIP):
        zip_path = DEFAULT_ZIP
    else:
        empty_df = pd.DataFrame(columns=["address_input", "CF", "lat", "lon"])
        return None, None, "", empty_df, "Shapefile の ZIP をアップロードするか、data/japan_ver85.zip を配置してください。"

    try:
        gdf_pref = load_gdf_from_zip(zip_path)
    except Exception as e:
        empty_df = pd.DataFrame(columns=["address_input", "CF", "lat", "lon"])
        return None, None, "", empty_df, f"行政界の読み込みに失敗しました: {e}"

    # Excel→ジオコーディング
    if excel_file is None or not hasattr(excel_file, "name"):
        gdf_pts = gpd.GeoDataFrame(columns=["address_input", "CF", "lat", "lon"], geometry=[], crs="EPSG:4326")
        table_df = pd.DataFrame(columns=["address_input", "CF", "lat", "lon"])
    else:
        try:
            df = pd.read_excel(excel_file.name, sheet_name=sheet_name, header=int(header_row))
        except Exception as e:
            empty_df = pd.DataFrame(columns=["address_input", "CF", "lat", "lon"])
            return None, None, "", empty_df, f"Excel の読み込みに失敗しました: {e}"

        addr_series = df.iloc[:, address_col] if isinstance(address_col, int) else df[address_col]
        cf_series = df.iloc[:, power_col] if isinstance(power_col, int) else df[power_col]

        addresses = addr_series.astype(str).tolist()
        cfs = cf_series.tolist()

        geo_df = geocode_with_cache(addresses, cfs, use_internet=bool(use_inet))
        table_df = geo_df[["address_input", "CF", "lat", "lon"]].copy()

        geometry = [
            Point(lon, lat) if (pd.notna(lat) and pd.notna(lon)) else None
            for lat, lon in zip(geo_df["lat"], geo_df["lon"])
        ]
        gdf_pts = gpd.GeoDataFrame(geo_df, geometry=geometry, crs="EPSG:4326")

    # 図と地図
    try:
        img = plot_map_png(
            gdf_pref, gdf_pts,
            line_width=float(line_width),
            marker_size=int(marker_size),
            legend_shrink=float(legend_shrink),
            legend_fontsize=int(legend_fontsize),
        )
    except Exception as e:
        return None, None, "", table_df, f"静的描画に失敗しました: {e}"

    try:
        html = make_folium_html(gdf_pref, gdf_pts, marker_size=int(marker_size))
    except Exception as e:
        html = f"<p>folium描画に失敗しました: {e}</p>"

    # 情報
    info = []
    info.append(f"都道府県レコード数: {len(gdf_pref)}")
    if gdf_pref.crs:
        info.append(f"PREF CRS: {gdf_pref.crs}")
    info.append(f"ポイント数(有効座標): {int(gdf_pts.geometry.notnull().sum())} / {len(gdf_pts)}")
    if not gdf_pts.empty and gdf_pts.crs:
        info.append(f"PTS CRS: {gdf_pts.crs}")

    return img, html, "\n".join(info), table_df, ""

# ----------------------------
# Gradio UI
# ----------------------------
with gr.Blocks(title="Japan Shapefile + Excel Geocoding Plotter (GSI)") as demo:
    gr.Markdown("## japan_ver85.shp(ZIP) + Excel住所 → 日本地図にプロット(凡例小・点大の調整可)")

    with gr.Row():
        zip_in = gr.File(label="Shapefile (ZIP)", file_count="single", file_types=[".zip"])
        xlsx_in = gr.File(label="Excelファイル(住所付き)", file_count="single", file_types=[".xlsx", ".xls"])

    with gr.Row():
        sheet = gr.Textbox(label="シート名", value="認定設備")
        header_row = gr.Number(label="ヘッダー行番号(0始まり)", value=2, precision=0)

    with gr.Row():
        address_col = gr.Textbox(label="住所列(列名 or 0始まり列番号)", value="発電設備の所在地")
        power_col = gr.Textbox(label="数値列(任意:列名 or 0始まり列番号)", value="発電出力(kW)")

    with gr.Row():
        use_inet = gr.Checkbox(label="国土地理院APIに問い合わせ(オフでキャッシュのみ使用)", value=True)
        line_width = gr.Slider(0.2, 2.0, value=0.6, step=0.1, label="境界線の太さ")

    # 見た目調整スライダ
    with gr.Row():
        marker_size = gr.Slider(4, 64, value=24, step=2, label="ポイントサイズ(matplotlib / folium)")
        legend_shrink = gr.Slider(0.3, 1.0, value=0.6, step=0.05, label="凡例の縮小率(小さいほど小さく)")
        legend_fontsize = gr.Slider(6, 16, value=8, step=1, label="凡例の目盛フォントサイズ")

    run_btn = gr.Button("描画")

    out_img = gr.Image(label="静的地図(matplotlib)", type="pil")
    out_html = gr.HTML(label="インタラクティブ地図(folium)")
    out_info = gr.Textbox(label="メタ情報", lines=4)
    out_table = gr.Dataframe(label="ジオコーディング結果(住所・緯度・経度・CF)", wrap=True)
    out_err = gr.Markdown(label="エラー", visible=True)

    def _parse(x):
        try:
            return int(x)
        except Exception:
            return x

    def app_run(zipf, xls, s, h, a, p, inet, lw, ms, lsh, lfs):
        return run(
            zipf, xls, s, int(h), _parse(a), _parse(p),
            inet, lw, ms, lsh, lfs
        )

    run_btn.click(
        fn=app_run,
        inputs=[zip_in, xlsx_in, sheet, header_row, address_col, power_col, use_inet, line_width, marker_size, legend_shrink, legend_fontsize],
        outputs=[out_img, out_html, out_info, out_table, out_err],
    )

if __name__ == "__main__":
    demo.launch()