df_hrrchy
|lefId |Lineage |
|-------|--------------------------------------|
|36326 |["36326","36465","36976","36091","82"]|
|36121 |["36121","36908","36976","36091","82"]|
|36380 |["36380","36465","36976","36091","82"]|
|36448 |["36448","36465","36976","36091","82"]|
|36683 |["36683","36465","36976","36091","82"]|
|36949 |["36949","36908","36976","36091","82"]|
|37349 |["37349","36908","36976","36091","82"]|
|37026 |["37026","36908","36976","36091","82"]|
|36879 |["36879","36465","36976","36091","82"]|
df_trans
|tranID | T_Id |
|-----------|-------------------------------------------------------------------------|
|1000540 |["36121","36326","37349","36949","36380","37026","36448","36683","36879"]|
df_creds
|T_Id |T_val |T_Goal |Parent_T_Id |Parent_Val |parent_Goal|
|-------|-------|-------|---------------|----------------|-----------|
|36448 |100 |1 |36465 |200 |1 |
|36465 |200 |1 |36976 |300 |2 |
|36326 |90 |1 |36465 |200 |1 |
|36091 |500 |19 |82 |600 |4 |
|36121 |90 |1 |36908 |200 |1 |
|36683 |90 |1 |36465 |200 |1 |
|36908 |200 |1 |36976 |300 |2 |
|36949 |90 |1 |36908 |200 |1 |
|36976 |300 |2 |36091 |500 |19 |
|37026 |90 |1 |36908 |200 |1 |
|37349 |100 |1 |36908 |200 |1 |
|36879 |90 |1 |36465 |200 |1 |
|36380 |90 |1 |36465 |200 |1 |
Desired Result
T_id
children
T_Val
T_Goal
parent_T_id
parent_Goal
trans_id
36091
["36976"]
500
19
82
4
1000540
36465
["36448","36326","36683","36879","36380"]
200
1
36976
2
1000540
36908
["36121","36949","37026","37349"]
200
1
36976
2
1000540
36976
["36465","36908"]
300
2
36091
19
1000540
36683
null
90
1
36465
1
1000540
37026
null
90
1
36908
1
1000540
36448
null
100
1
36465
1
1000540
36949
null
90
1
36908
1
1000540
36326
null
90
1
36465
1
1000540
36380
null
90
1
36465
1
1000540
36879
null
90
1
36465
1
1000540
36121
null
90
1
36908
1
1000540
37349
null
100
1
36908
1
1000540
Code Tried
from pyspark.sql import functions as F
from pyspark.sql import DataFrame
from pyspark.sql.functions import explode, collect_set, expr, col, collect_list,array_contains, lit
from functools import reduce
for row in df_transactions.rdd.toLocalIterator():
# def find_nodemap(row):
dfs = []
df_hy_set = (df_hrrchy.filter(df_hrrchy. lefId.isin(row["T_ds"]))
.select(explode("Lineage").alias("Terrs"))
.agg(collect_set(col("Terrs")).alias("hierarchy_list"))
.select(F.lit(row["trans_id"]).alias("trans_id "),"hierarchy_list")
)
df_childrens = (df_creds.join(df_ hy _set, expr("array_contains(hierarchy_list, T_id)"))
.select("T_id", "T_Val","T_Goal","parent_T_id", "parent_Goal", "trans _id" )
.groupBy("parent_T_id").agg(collect_list("T_id").alias("children"))
)
df_filter_creds = (df_creds.join(df_ hy _set, expr("array_contains(hierarchy_list, T_id)"))
.select ("T_id", "T_val","T_Goal","parent_T_id", "parent_Goal”, "trans_id")
)
df_nodemap = (df_filter_ creds.alias("A").join(df_childrens.alias("B"), col("A.T_id") == col("B.parent_T_id"), "left")
.select("A.T_id","B.children", "A.T_val","A.terr_Goal","A.parent_T_id", "A.parent_Goal", "A.trans_ id")
)
display(df_nodemap)
# dfs.append(df_nodemap)
# df = reduce(DataFrame.union, dfs)
# display(df)
# # display(df)
My problem - Its a bad design. df_trans is having millions of data and looping through dataframe , its taking forever. Without looping can I do it. I tried couple of other methods, not able to get the desired result.
You certainly need to process entire DataFrame in batch, not iterate row by row.
Key points are to "reverse" df_hrrchy, ie. from parent lineage obtain list of children for every T_Id:
val df_children = df_hrrchy.withColumn("children", slice($"Lineage", lit(1), size($"Lineage") - 1))
.withColumn("parents", slice($"Lineage", 2, 999999))
.select(explode(arrays_zip($"children", $"parents")).as("rels"))
.distinct
.groupBy($"rels.parents".as("T_Id"))
.agg(collect_set($"rels.children").as("children"))
df_children.show(false)
+-----+-----------------------------------+
|T_Id |children |
+-----+-----------------------------------+
|36091|[36976] |
|36465|[36448, 36380, 36326, 36879, 36683]|
|36976|[36465, 36908] |
|82 |[36091] |
|36908|[36949, 37349, 36121, 37026] |
+-----+-----------------------------------+
then expand list of T_Ids in df_trans and also include all T_Ids from the hierarchy:
val df_trans_map = df_trans.withColumn("T_Id", explode($"T_Id"))
.join(df_hrrchy, array_contains($"Lineage", $"T_Id"))
.select($"tranID", explode($"Lineage").as("T_Id"))
.distinct
df_trans_map.show(false)
+-------+-----+
|tranID |T_Id |
+-------+-----+
|1000540|36976|
|1000540|82 |
|1000540|36091|
|1000540|36465|
|1000540|36326|
|1000540|36121|
|1000540|36908|
|1000540|36380|
|1000540|36448|
|1000540|36683|
|1000540|36949|
|1000540|37349|
|1000540|37026|
|1000540|36879|
+-------+-----+
With this it is just a simple join to obtain final result:
df_trans_map.join(df_creds, Seq("T_Id"))
.join(df_children, Seq("T_Id"), "left_outer")
.show(false)
+-----+-------+-----+------+-----------+----------+-----------+-----------------------------------+
|T_Id |tranID |T_val|T_Goal|Parent_T_Id|Parent_Val|parent_Goal|children |
+-----+-------+-----+------+-----------+----------+-----------+-----------------------------------+
|36976|1000540|300 |2 |36091 |500 |19 |[36465, 36908] |
|36091|1000540|500 |19 |82 |600 |4 |[36976] |
|36465|1000540|200 |1 |36976 |300 |2 |[36448, 36380, 36326, 36879, 36683]|
|36326|1000540|90 |1 |36465 |200 |1 |null |
|36121|1000540|90 |1 |36908 |200 |1 |null |
|36908|1000540|200 |1 |36976 |300 |2 |[36949, 37349, 36121, 37026] |
|36380|1000540|90 |1 |36465 |200 |1 |null |
|36448|1000540|100 |1 |36465 |200 |1 |null |
|36683|1000540|90 |1 |36465 |200 |1 |null |
|36949|1000540|90 |1 |36908 |200 |1 |null |
|37349|1000540|100 |1 |36908 |200 |1 |null |
|37026|1000540|90 |1 |36908 |200 |1 |null |
|36879|1000540|90 |1 |36465 |200 |1 |null |
+-----+-------+-----+------+-----------+----------+-----------+-----------------------------------+
You need to re-write this to use the full cluster, using a localIterator means that you aren't fully utilizing the cluster for shared work.
Below code was not run as you didn't provide a workable data set to test. If you do I'll run the code to make sure it's sound.
from pyspark.sql import functions as F
from pyspark.sql import DataFrame
from pyspark.sql.functions import explode, collect_set, expr, col, collect_list,array_contains, lit
from functools import reduce
#uses explode I know this will create a lot of short lived records but the flip side is it will use the entire cluster to complete the work instead of the driver.
df_trans_expld = df_trans.select( df_trans.tranID, explode(df_trans.T_Id).alias("T_Id") )
#uses explode
df_hrrchy_expld = df_hrrchy.select( df_hrrchy.leftId, explode( df_hrrchy.Lineage ).alias("Lineage") )
#uses exploded data to join which is the same as a filter.
df_hy_set = df_trans_expld.join( df_hrrchy_expld, df_hrrchy_expld.lefId === df_trans_expld.T_id, "left").select( "trans_id" ).agg(collect_set(col("Lineage")).alias("hierarchy_list"))
.select(F.lit(col("trans_id")).alias("trans_id "),"hierarchy_list")
#logic unchanged from here down
df_childrens = (df_creds.join(df_hy_set, expr("array_contains(hierarchy_list, T_id)"))
.select("T_id", "T_Val","T_Goal","parent_T_id", "parent_Goal", "trans _id" )
.groupBy("parent_T_id").agg(collect_list("T_id").alias("children"))
)
df_filter_creds = (df_creds.join(ddf_hy_set, expr("array_contains(hierarchy_list, T_id)"))
.select ("T_id", "T_val","T_Goal","parent_T_id", "parent_Goal”, "trans_id")
)
df_nodemap = (df_filter_creds.alias("A").join(df_childrens.alias("B"), col("A.T_id") == col("B.parent_T_id"), "left")
.select("A.T_id","B.children", "A.T_val","A.terr_Goal","A.parent_T_id", "A.parent_Goal", "A.trans_ id")
)
# no need to append/union data as it's now just one dataframe df_nodemap
I'd have to look into this more but I'm pretty sure you are pulling all the data through the driver(with your existing code), which will really slow things down, this will make use of all executors to complete the work.
There may be another optimization to get rid of the array_contains (and use a join instead). I'd have to look at the explain to see if you could get even more performance out of it. Don't remember off the top of my head, you are avoiding a shuffle so it may be better as is.
I have a dataframe with data like this
I want to compare each record with next year's record and see if that id is there. If it's there in the year and the next year it's 'Existing', If it's there in the year and not there in the next year, it's 'Left'. If it's not there in the year but there in the next year, it's 'New'. I want output like this below. The columns 2017-18, 2018-19 etc. should be created dynamically.
How do I achieve this?
After getting the data in the above format, I need to aggregate the sales for each year band like this as below. For example for 2017-2018,
New_sales = sum of all sales of 2018 (which is the later year in 2017-2018) where it's marked as 'New' which is 25 here.
Left_sales = sum of all sales of 2017 (the earlier year in 2017-2018) where it's marked as 'Left' which is 100 here.
Existing_sales = sum of sales of 2017 where it's marked as 'Existing' subtract sum of sales of 2018 where it's marked as 'Existing'
Existing_sales = 50+75 (sales of 2017, 'Existing') - (20+50) (sales of 2018, 'Existing') = 125-70 = 55
How do I achieve this?
As your date is a string, I think you can:
df = spark.createDataFrame(
data=[
(1, '31/12/2017'),
(2, '31/12/2017'),
(3, '31/12/2017'),
(1, '31/12/2018'),
(3, '31/12/2018'),
(5, '31/12/2018'),
],
schema=['id', 'date']
)
First you can get the year first:
df2 = df.withColumn('year', func.split(func.col('Date'), '/').getItem(2))
df2.show(10, False)
+---+----------+----+
|id |date |year|
+---+----------+----+
|1 |31/12/2017|2017|
|2 |31/12/2017|2017|
|3 |31/12/2017|2017|
|1 |31/12/2018|2018|
|3 |31/12/2018|2018|
|5 |31/12/2018|2018|
+---+----------+----+
Then you can collect the list of year by id as a reference:
df3 = df2.groupby('id')\
.agg(func.collect_set('year').alias('year_lst'))
df3.show(3, False)
+---+------------+
|id |year_lst |
+---+------------+
|1 |[2017, 2018]|
|2 |[2017] |
|3 |[2017, 2018]|
+---+------------+
Then you can join the reference back to the data:
df4 = df2.join(df3, on='id', how='left')
df4.show(10, False)
+---+----------+----+------------+
|id |date |year|year_lst |
+---+----------+----+------------+
|1 |31/12/2017|2017|[2017, 2018]|
|2 |31/12/2017|2017|[2017] |
|3 |31/12/2017|2017|[2017, 2018]|
|1 |31/12/2018|2018|[2017, 2018]|
|3 |31/12/2018|2018|[2017, 2018]|
|5 |31/12/2018|2018|[2018] |
+---+----------+----+------------+
The last step is to create the column dynamically. I think you can use a for loop:
year_loop = ['2017', '2018', '2019', '2020', '2021']
for idx in range(len(year_loop)-1):
this_year = year_loop[idx]
next_year = year_loop[idx+1]
column_name = f"{this_year}-{next_year}"
new_condition = (~func.array_contains(func.col('year_lst'), this_year)) & (func.array_contains(func.col('year_lst'), next_year))
exist_condition = (func.array_contains(func.col('year_lst'), this_year)) & (func.array_contains(func.col('year_lst'), next_year))
left_condition = (func.array_contains(func.col('year_lst'), this_year)) & (~func.array_contains(func.col('year_lst'), next_year))
df4 = df4.withColumn(column_name, func.when(new_condition, func.lit('New'))
.when(exist_condition, func.lit('Existing'))
.when(left_condition, func.lit('Left')))
df4.show(10, False)
+---+----------+----+------------+---------+---------+---------+---------+
|id |date |year|year_lst |2017-2018|2018-2019|2019-2020|2020-2021|
+---+----------+----+------------+---------+---------+---------+---------+
|1 |31/12/2017|2017|[2017, 2018]|Existing |Left |null |null |
|2 |31/12/2017|2017|[2017] |Left |null |null |null |
|3 |31/12/2017|2017|[2017, 2018]|Existing |Left |null |null |
|1 |31/12/2018|2018|[2017, 2018]|Existing |Left |null |null |
|3 |31/12/2018|2018|[2017, 2018]|Existing |Left |null |null |
|5 |31/12/2018|2018|[2018] |New |Left |null |null |
+---+----------+----+------------+---------+---------+---------+---------+
I need to filter a dataframe with the below criteria.
I have 2 columns 4Wheel(Subaru, Toyota, GM, null/empty) and 2Wheel(Yamaha, Harley, Indian, null/empty).
I have to filter on 4Wheel with values (Subaru, Toyota), if 4Wheel contain empty/null then filter on 2Wheel with values (Yamaha, Harley)
I couldn't find this type of filtering in different examples. I am new to spark/scala, so could not get enough idea to implement this.
Thanks,
Barun.
You can use spark SQL built-in function when to check if a column is null or empty, and filter accordingly:
import org.apache.spark.sql.functions.{col, when}
dataframe.filter(when(col("4Wheel").isNull || col("4Wheel").equalTo(""),
col("2Wheel").isin("Yamaha", "Harley")
).otherwise(
col("4Wheel").isin("Subaru", "Toyota")
))
So if you have the following input:
+---+------+------+
|id |4Wheel|2Wheel|
+---+------+------+
|1 |Toyota|null |
|2 |Subaru|null |
|3 |GM |null |
|4 |null |Yamaha|
|5 | |Yamaha|
|6 |null |Harley|
|7 | |Harley|
|8 |null |Indian|
|9 | |Indian|
|10 |null |null |
+---+------+------+
You get the following filtered ouput:
+---+------+------+
|id |4Wheel|2Wheel|
+---+------+------+
|1 |Toyota|null |
|2 |Subaru|null |
|4 |null |Yamaha|
|5 | |Yamaha|
|6 |null |Harley|
|7 | |Harley|
+---+------+------+
I obtain a resultant dataframe after performing some computations over it.Say the dataframe is result. When i write it to Amazon S3 there are specific cells which are shown blank. The top 5 of my result dataframe is:
_________________________________________________________
|var30 |var31 |var32 |var33 |var34 |var35 |var36|
--------------------------------------------------------
|-0.00586|0.13821 |0 | |1 | | |
|3.87635 |2.86702 |2.51963 |8 |11 |2 |14 |
|3.78279 |2.54833 |2.45881 | |2 | | |
|-0.10092|0 |0 |1 |1 |3 |1 |
|8.08797 |6.14486 |5.25718 | |5 | | |
---------------------------------------------------------
But when i run result.show() command i am able to see the values.
_________________________________________________________
|var30 |var31 |var32 |var33 |var34 |var35 |var36|
--------------------------------------------------------
|-0.00586|0.13821 |0 |2 |1 |1 |6 |
|3.87635 |2.86702 |2.51963 |8 |11 |2 |14 |
|3.78279 |2.54833 |2.45881 |2 |2 |2 |12 |
|-0.10092|0 |0 |1 |1 |3 |1 |
|8.08797 |6.14486 |5.25718 |20 |5 |5 |34 |
---------------------------------------------------------
Also, the blank are shown in same cells every time i run it.
Use this to save data to your s3
DataFrame.repartition(1).write.format("com.databricks.spark.csv").option("header", "true").save("s3n://Yourpath")
For anyone who might have come across this issue, I can tell what worked for me.
I was joining 1 data frame ( let's say inputDF) with another df ( delta DF) based on some logic and storing in an output data frame (outDF). I was getting same error where by I could see a record in outDF.show() but while writing this dataFrame into a hive table OR persisting the outDF ( using outDF.persist(StorageLevel.MEMORY_AND_DISC)) I wasn't able to see that particular record.
SOLUTION:- I persisted the inputDF ( inputDF.persist(StorageLevel.MEMORY_AND_DISC)) before joining it with deltaDF. After that outDF.show() output was consistent with the hive table where outDF was written.
P.S:- I am not sure how this solved the issue. Would be awesome if someone could explain this, but the above worked for me.