global impact,loans,countries,members

about = pd.read_csv("about.csv",sep=",")

impact = pd.read_csv("impact.csv",sep=",")

loans = pd.read_csv("loans.csv",sep=",")

countries = pd.read_csv("countries.csv",sep=",")

members = pd.read_csv("members.csv",sep=",")

impact = impact.dropna(how="any")

intro = about[["team_url","intro"]]

impact = pd.merge(impact,intro,how="inner",on="team_url")

impact = impact.fillna("")

impact["intro_len"] = impact["intro"].str.split().apply(len)

impact["loans_per_member"] = impact["loans_per_member"].astype(float)

impact["total_loans"] = impact["total_loans"].astype(int)

impact["total_amount_lent"] = impact["total_amount_lent"].astype(float)

impact["borrower_female_count"] = impact["borrower_female_count"].astype(int)

impact["borrower_male_count"] = impact["borrower_male_count"].astype(int)

global loans_per_member, total_loans, total_amount_lent,borrower_female_count,borrower_male_count

loans_per_member = impact["loans_per_member"]

total_loans = impact["total_loans"]

total_amount_lent = impact["total_amount_lent"]

borrower_female_count = impact["borrower_female_count"]

print "Minimum Values"

print "Min of loans per members: {0:0.2f}".format(loans_per_member.min())

print "Min of total loans: {0}".format(total_loans.min())

print "Min of total amount_lent: {0:0.2f}".format(total_amount_lent.min())

print "Min of borrower male count: {0}".format(borrower_male_count.min())

print "Maximum Values"

print "Max of loans per members: {0:0.2f}".format(loans_per_member.max())

print "Max of total loans: {0}".format(total_loans.max())

print "Max of total amount_lent: {0:0.2f}".format(total_amount_lent.max())

print "Max of borrower male count: {0}".format(borrower_male_count.max())

mode = impact.mode(numeric_only=True)

print "Mode Values"

print "Mode of loans per member: ",mode["loans_per_member"].loc[0]

print "Mode of total loans: ",mode["total_loans"].loc[0]

print "Mode of total amount lent: ", mode["total_amount_lent"].loc[0]

print "Mode of borrower male count: ", mode["borrower_male_count"].loc[0]

print "Median Values"

print "Median of loans per members: {0:0.2f}".format(loans_per_member.median())

print "Median of total loans: {0}".format(total_loans.median())

print "Median of total amount_lent: {0:0.2f}".format(total_amount_lent.median())

print "Median of borrower male count: {0}".format(borrower_male_count.median())

print "Mean Values"

print "Mean of loans per members: {0:0.2f}".format(loans_per_member.mean())

print "Mean of total loans: {0:0.2f}".format(total_loans.mean())

print "Mean of total amount_lent: {0:0.2f}".format(total_amount_lent.mean())

print "Mean of borrower male count: {0:0.2f}".format(borrower_male_count.mean())

print "Standard Deviation Values"

print "Std of loans per members: {0:0.2f}".format(loans_per_member.std())

print "Std of total loans: {0:0.2f}".format(total_loans.std())

print "Std of total amount lent: {0:0.2f}".format(total_amount_lent.std())

print "Std of borrower male count: {0:0.2f}".format(borrower_male_count.std())

# outlier is defined as outside of 2 std deviation

print "Outliers values"

print "Number of loans per member"

out = getOutliers(loans_per_member)

print "Number of loans per member outliers: ", out.count()

print "Number of total loans"

out = getOutliers(total_loans)

print "Number of total loans outliers: ", out.count()

print "Number of total amount lent"

out = getOutliers(total_amount_lent)

print "Number of total amount lent outliers: ", out.count()

print "Number of total borrower male count"

out = getOutliers(borrower_male_count)

print "Number of total borrower male count outliers: ", out.count()

print "Number of total borrower female count"

out = getOutliers(borrower_female_count)

boolean1 = df < df.mean()-2*df.std()

boolean2 = df > df.mean()+2*df.std()

boolean = boolean1 | boolean2

print impact[boolean]["team_url"]

plot1a = histoPlot(title="Loans per member (Untrimmed)",s=loans_per_member,div=50,xlabel="Loans per member")

plot1b = histoPlot(title="Loans per member (Trimmed)",s=getTrimmed(loans_per_member),div=200,xlabel="Loans per member")

count,division = np.histogram(getTrimmed(loans_per_member),bins = 200)

print "LPM: divisions {0} to {1} has the maximum count of {2}".format(division[0],division[1],count[0])

plot1a.savefig("figures/Loans per member (untrimemd).png")

plot1b.savefig("figures/Loans per member (trimmed).png")

plot2a = histoPlot(title="Total loans (Untrimmed)", s=total_loans,div = 50, xlabel="Total loans")

plot2b = histoPlot(title="Total loans (Trimmed)", s = getTrimmed(total_loans), div = 200, xlabel="Total loans")

count,division = np.histogram(getTrimmed(total_loans),bins = 200)

print "TL: divisions {0} to {1} has the maximum count of {2}".format(division[0],division[1],count[0])

plot2a.savefig("figures/Total loans (untrimmed).png")

plot2b.savefig("figures/Total loans (trimmed).png")

plot3a = histoPlot(title="Total amount lent (Untrimmed)", s=total_amount_lent,div = 50, xlabel="Total amount lent")

plot3b = histoPlot(title="Total amount lent (Trimmed)", s = getTrimmed(total_amount_lent), div = 200, xlabel="Total amount lent")

count,division = np.histogram(getTrimmed(total_amount_lent),bins = 200)

print "TAL: divisions {0} to {1} has the maximum count of {2}".format(division[0],division[1],count[0])

plot3a.savefig("figures/Total amount lent (untrimmed).png")

plot3b.savefig("figures/Total amount lent (trimmed).png")

plot4a = histoPlot(title="Borrower male count (Untrimmed)", s=borrower_male_count,div = 50, xlabel="Borrower male count")

plot4b = histoPlot(title="Borrower male count (Trimmed)", s = getTrimmed(borrower_male_count), div = 200, xlabel="Borrower male count")

count,division = np.histogram(getTrimmed(borrower_male_count),bins = 200)

print "BMC: divisions {0} to {1} has the maximum count of {2}".format(division[0],division[1],count[0])

plot4a.savefig("figures/Borrower male count (untrimmed).png")

plot4b.savefig("figures/Borrower male count (trimmed).png")

plot5a = histoPlot(title="Borrower female count (Untrimmed)", s=borrower_female_count,div = 50, xlabel="Borrower female count")

plot5b = histoPlot(title="Borrower female count (Trimmed)", s = getTrimmed(borrower_female_count), div = 200, xlabel="Borrower female count")

count,division = np.histogram(getTrimmed(borrower_female_count),bins = 200)

print "BFC: divisions {0} to {1} has the maximum count of {2}".format(division[0],division[1],count[0])

plot5a.savefig("figures/Borrower female count (untrimmed).png")

plot5b.savefig("figures/Borrower female count (trimmed).png")

plot6a = histoPlot(title="Length of introduction (Untrimmed)", s = impact["intro_len"],div = 50, xlabel="Length of introduction")

plot6b = histoPlot(title="Length of introduction (Trimmed)",s=impact["intro_len"],div = 200, xlabel = "Length of introduction")

plot6a.savefig("figures/Length of introduction (untrimmed).png")

fig = plt.figure()

s.hist(bins=div,log=True)

plt.title(title)

plt.xlabel(xlabel)

plt.ylabel("Frequency")

merged = pd.concat([total_amount_lent,total_loans, loans_per_member, borrower_female_count,borrower_male_count],axis = 1)

result = sm.ols(formula="total_amount_lent ~ total_loans", data = merged).fit()

print result.summary()

result = sm.ols(formula="total_amount_lent ~ loans_per_member", data = merged).fit()

print result.summary()

result = sm.ols(formula="total_amount_lent ~ borrower_female_count", data = merged).fit()

print result.summary()

result = sm.ols(formula="total_amount_lent ~ borrower_male_count", data = merged).fit()

print result.summary()

result = sm.ols(formula = "total_amount_lent ~ intro_len", data = impact).fit()

plot1 = scatterPlot(title="Total amount lent vs. total loans",x=total_loans,y=total_amount_lent,

xlabel="Total number of loans",ylabel="Total amount lent ($)")

plot2 = scatterPlot(title="Total amount lent vs. loans per member",x=loans_per_member,y=total_amount_lent,

xlabel="Loans per member",ylabel="Total amount lent ($)")

plot3 = scatterPlot(title="Total amount lent vs. number of female borrowers",x=borrower_female_count,y=total_amount_lent,

xlabel="number of female borrowers",ylabel="Total amount lent ($)")

plot4 = scatterPlot(title="Total amount lent vs. number of male borrowers",x=borrower_male_count,y=total_amount_lent,

xlabel="number of male borrowers",ylabel="Total amount lent ($)")

plot4 = scatterPlot(title="Total amount lent vs. number of male borrowers",x=borrower_male_count,y=total_amount_lent,

xlabel="number of male borrowers",ylabel="Total amount lent ($)")

plot5 = scatterPlot(title="Total amount lent vs. length of introduction", x=impact["intro_len"],y = total_amount_lent,

xlabel = "length of introduction", ylabel = "Total amount lent ($)")

plot1.savefig("figures/Total amount lent vs. total loans.png")

plot2.savefig("figures/Total amount lent vs. loans per member.png")

plot3.savefig("figures/Total amount lent vs. number of female borrowers.png")

plot4.savefig("figures/Total amount lent vs. number of male borrowers.png")

fig = plt.figure()

plt.scatter(x=x,y=y)

plt.title(title)

plt.xlabel(xlabel)

plt.ylabel(ylabel)

m,b = np.polyfit(x,y,1)

plt.plot(x,m*x+b,"-")

n = len(data)

sum_xy = 0

sum_x =0

sum_y = 0

sum_x_sq = 0

m = 0

for x,y in data:

sum_xy = sum_xy + x*y

sum_x = sum_x + x

sum_y = sum_y + y

sum_x_sq = sum_x_sq + x**2

m = float((n*sum_xy - sum_x * sum_y)/(n*sum_x_sq - sum_x**2))

b = float((sum_y - m*sum_x)/n)