#!/bin/env python3
"""
Extract area of cancerous tissue (larger than 200um^2)
and extract the percentage area of cancerous tissue per square
"""
import pandas as pd
import numpy as np
from scipy import stats


################################################################################
# SW620
################################################################################
meta = pd.read_csv("../data/20220406_metadata_micrometastasis_quantification.csv")
measure = pd.read_csv("../data/measurements_29_april_237pm.tsv", sep="\t")

measure["Image_crop"] = measure["Image"].apply(lambda x: x.split(".vsi")[0])
merge = pd.merge(measure, meta, left_on="Image_crop", right_on="Image_name")
merge_filt = merge.loc[merge["Area µm^2"] >= 200]

total_area = pd.read_csv("../data/measurements_area_squares.tsv", sep="\t")
total_area = total_area.loc[total_area["ROI"] == "Rectangle", ["Image", "Area µm^2"]].drop_duplicates()
total_area.columns = ["Image", "Total Area"]

sel_cancer = merge_filt["Class"] == "cancer_cells"
merge_cancer = merge_filt.loc[sel_cancer == True]

# Area
meta["animal"] = meta["Identifier"].apply(lambda x: x.split("_")[2])
squares_area = merge_cancer.groupby(["Image", "Parent"])["Area µm^2"].sum().reset_index()
cond = merge_cancer.groupby(["Image", "Parent"])["Condition"].first().reset_index()
animal = merge_cancer.groupby(["Image", "Parent"])["Condition"].first().reset_index()
squares_area = pd.merge(squares_area, cond)
squares_area = pd.merge(squares_area, total_area)
squares_area["Image_crop"] = squares_area["Image"].apply(lambda x: x.split(".vsi")[0])
squares_area = pd.merge(squares_area, meta[["Image_name", "animal"]], left_on="Image_crop",
        right_on="Image_name")
squares_area["percent area"] = squares_area["Area µm^2"] / squares_area["Total Area"] * 100
squares_area = squares_area.reset_index()

# Remove misassigned square
s = (squares_area["percent area"] > 3) & (squares_area["Condition"] == "GFP")
squares_area = squares_area.loc[~s]

squares_area.to_csv("../results/sw620_squares_area.csv")

pou_area = squares_area.loc[squares_area["Condition"] == "POU5F1B", "percent area"]
gfp_area = squares_area.loc[squares_area["Condition"] == "GFP", "percent area"]

t, pval = stats.ttest_ind(pou_area, gfp_area)

################################################################################
# LS1034
################################################################################
meta = pd.read_csv("../data/20220504_metadata_micrometastasis_quantification_LS1034_knockdown_slide_scanner.csv")
measure = pd.read_csv("../data/20220504_detection_measurements_micrometastasis_quantification_LS1034_knockdown_slide_scanner.tsv", sep="\t")
measure["Image_crop"] = measure["Image"].apply(lambda x: x.split(".vsi")[0])

merge = pd.merge(measure, meta, left_on="Image_crop", right_on="Image_name")
merge_filt = merge.loc[merge["Area µm^2"] >= 200]

total_area = pd.read_csv("../data/20220504_annotation_measurements_micrometastasis_quantification_LS1034_knockdown_slide_scanner.tsv", sep="\t")
total_area = total_area.loc[total_area["ROI"] == "Rectangle", ["Image", "Area µm^2"]].drop_duplicates()
total_area.columns = ["Image", "Total Area"]

sel_cancer = merge_filt["Class"] == "cancer_cells"
merge_cancer = merge_filt.loc[sel_cancer == True]

# Area
meta["animal"] = meta["Identifier"].apply(lambda x: "-".join(x.split("_")[1:3]))
squares_area = merge_cancer.groupby(["Image", "Parent"])["Area µm^2"].sum().reset_index()
cond = merge_cancer.groupby(["Image", "Parent"])["Condition"].first().reset_index()
animal = merge_cancer.groupby(["Image", "Parent"])["Condition"].first().reset_index()
squares_area = pd.merge(squares_area, cond)
squares_area = pd.merge(squares_area, total_area)
squares_area["Image_crop"] = squares_area["Image"].apply(lambda x: x.split(".vsi")[0])
squares_area = pd.merge(squares_area, meta[["Image_name", "animal"]], left_on="Image_crop",
        right_on="Image_name")
squares_area["percent area"] = squares_area["Area µm^2"] / squares_area["Total Area"] * 100
squares_area = squares_area.reset_index()

squares_area.to_csv("../results/ls1034_squares_area.csv")

sh_area = squares_area.loc[squares_area["Condition"] == "sh3", "percent area"]
nts_area = squares_area.loc[squares_area["Condition"] == "NTS", "percent area"]

t, pval = stats.ttest_ind(sh_area, nts_area)