import pandas as pd
import numpy as np
import geopandas as gpd

import requests

url = "https://Projects-by-IF-model-temp-api.hf.space/model-effect"
headers = {"Content-Type": "application/json"}

cell_area = 100*100 #in square meters

def make_XYT(df, treatment_col = '%CoberturaVeg', target_col = 'LST'):
    
    Y = df[target_col].values

    T = df[treatment_col].values

    X = df.drop(columns=[treatment_col, target_col]).values
    
    
    return X, Y, T
    
def prepare_data(og_file, training_file):
    original_df = pd.read_csv(og_file, index_col=0)
    df = pd.read_csv(training_file, index_col=0)

    original_df = original_df.round(2)
    
    df = df.drop(columns=['row','column','x','y'])
    df = df.astype(np.float32)

    X, Y, T = make_XYT(df)
    
    return original_df, df, X, Y, T


def prepare_treatment_df(original_df, df):
    treatment_df = original_df[['x','y','%CoberturaVeg']].copy()
    treatment_df.loc[:, '%Construccion'] = original_df['%Construccion'].values
    treatment_df.loc[:, 'LST'] = df['LST'].values
    
    treatment_df = treatment_df.reset_index()
    return treatment_df
    

def run_treatment_increase(X, T, original_df, df, value, name):
    
    treatment_df = prepare_treatment_df(original_df, df)
    T_sim = T.copy()  
    value_num = 1+ (value/100)
    T_sim = T_sim*value_num
    data = {"X": X.tolist(), "T0":T.tolist(), "T1":(T_sim).tolist()}
    response = requests.post(url, headers=headers, json=data)
    if response.ok:
        result = response.json()
        # print("Model response:", result["effect"])
    else:
        print("Request failed:", response.status_code, response.text)
    
    treatment_df[name] = result["effect"]
    treatment_df[f'%total_trees_treatment_{value}%'] = T_sim
    treatment_df[f'pp_trees_increase_treatment_{value}%'] = treatment_df[f'%total_trees_treatment_{value}%'] - treatment_df['%CoberturaVeg']

    return treatment_df

def process_data_trees_to_temp(value, original_df, df, X, Y, T):

    name = f'treatment_effect_{value}%'
    treatment_df = run_treatment_increase(X, T, original_df, df, value, name)

    treatment_df[f'simulated_temp_{value}%'] = treatment_df['LST'] + treatment_df[name]

    #add columns translated to square meters
    treatment_df[f'sqm_total_trees_treatment_{value}%'] = (treatment_df[f'%total_trees_treatment_{value}%']/100)*cell_area
    treatment_df[f'sqm_trees_increase_treatment_{value}%'] = (treatment_df[f'pp_trees_increase_treatment_{value}%']/100)*cell_area

    gdf = gpd.GeoDataFrame(treatment_df, geometry=gpd.points_from_xy(treatment_df.x, treatment_df.y))

    return gdf


def run_temp_decrease(X, original_df, df, goal, name):
    
    treatment_df = prepare_treatment_df(original_df, df)
    data = {"X": X.tolist()}
    response = requests.post(url, headers=headers, json=data)
    if response.ok:
        result = response.json()
        # print("Model response:", result["effect"])
    else:
        print("Request failed:", response.status_code, response.text)
    
    treatment_effects = np.array(result["effect"])
    
    trees_needed = -goal / treatment_effects
    
    treatment_df[name] = trees_needed
    treatment_df['goal_temp'] = treatment_df['LST'] - goal

    return treatment_df
def process_data_temp_to_trees(goal, original_df, df, X, Y, T):

    name = f'pp_trees_increase_for_{goal}C'
    treatment_df = run_temp_decrease(X,original_df, df, goal, name)

    treatment_df[f'%total_trees_needed_for_{goal}C'] = treatment_df['%CoberturaVeg'] + treatment_df[name]

    #add columns translated to square meters
    treatment_df[f'sqm_total_trees_needed_for_{goal}C'] = (treatment_df[f'%total_trees_needed_for_{goal}C']/100)*cell_area
    treatment_df[f'sqm_trees_increase_for_{goal}C'] = (treatment_df[f'pp_trees_increase_for_{goal}C']/100)*cell_area

    gdf = gpd.GeoDataFrame(treatment_df, geometry=gpd.points_from_xy(treatment_df.x, treatment_df.y))

    return gdf